Height-based muscle criteria help predict survival in cirrhosis
A survival prediction measure improved by 0.033 with the 2025 height-based definition, versus 0.023 with the 2019 definition.
Our read
Biomarkers- Studied:
- Cohort study of 505 people aged 50 or older with cirrhosis, median 5.4 years
- What changed:
- Adding height-based muscle criteria improved the survival prediction score (C-index) by 0.033 over the baseline model
- The catch:
- The study did not measure survival benefits from treating muscle loss
- Why it’s aging research:
- It measures an outcome that worsens with age.
Editor of record: Boris Djordjevic · AgingResearch.News
In a retrospective cohort study, researchers analyzed 505 people aged 50 or older with liver cirrhosis, or severe liver scarring. Participants underwent tests using electrical signals to measure body composition between 2016 and 2025.
The researchers compared three definitions from the Asian Working Group for Sarcopenia. Sarcopenia is a condition involving low muscle mass and reduced muscle function. Two definitions adjusted muscle mass for height; the third, added in 2025, used body mass index, a weight-to-height measure. During a median follow-up of 5.4 years, 140 people died. All three definitions were independently associated with mortality. The 2025 height-based definition predicted survival comparably to or slightly better than the 2019 version. Adding the body-mass-index definition to a model using age, sex, cause of liver disease and a liver function score provided little extra predictive value.
Why it matters
For adults aged 50 or older with cirrhosis, the study addressed how muscle assessments might inform estimates of survival.
Caveats
This retrospective observational study assessed prediction, not whether muscle loss caused deaths or whether treating it would improve survival. The findings concern people aged 50 or older with cirrhosis, rather than younger people or those without the disease.
The paper
Iwate Medical University
Hepatology International, Jul 15, 2026, Open access, accepted manuscript
Comments
Be the first to comment.