Overlap between body composition abnormalities and sex-specific prognostication in decompensated cirrhosis

TL;DR

This study explores how body composition differences affect survival in cirrhosis patients, finding distinct patterns between men and women.

Contribution

The study identifies sex-specific body composition profiles and their combined impact on mortality in decompensated cirrhosis patients.

Findings

High visceral adiposity is the most prevalent body composition abnormality in both male and female patients.

Overlapping body composition abnormalities are strongly associated with increased mortality risk in female patients.

A nomogram integrating body composition and clinical scores improves prognostication accuracy.

Abstract

We aimed to demonstrate distinct body composition (BC) profiles stratified by sex and clarify their joint effects on long-term mortality in a retrospective cohort of inpatients.

Various BC parameters annotated on computed tomography (CT) images at the third lumbar vertebra were used to define sarcopenia, myosteatosis, low subcutaneous adiposity, and high visceral adiposity. These categories were constructed using sex-specific, outcome-based cutoffs in a prerequisite manner.

Among 519 patients hospitalized for acute decompensating episodes, the median age was 64.0 years, with a slight female predominance (51.6%). Among the female patients, high visceral adiposity was the most prevalent single BC abnormality (38.4%), while the most common overlapping phenotype was myosteatosis occurring concurrently with high visceral adiposity (9.7%). Among the male patients, high visceral adiposity also showed the highest prevalence (74.9%), while the most common overlapping phenotype was sarcopenia occurring concurrently with low subcutaneous adiposity (15.1%). Considering their jointly negative impact, the female patients experiencing three BC abnormalities had the lowest survival rate (33.3%, log-rank test: p = 0.0022). Still, this difference was only marginally significant in the male patients with three or more BC abnormalities (log-rank test: p = 0.068). Furthermore, overlapped BC abnormalities were associated with 722 and 331% higher risks, respectively, of 1-year all-cause mortality (p = 0.001) in the female patients relative to those with no BC abnormalities and those with an isolated BC abnormality. Lastly, our established nomogram integrated albumin, Model for End-Stage Liver Disease-Sodium (MELD-Na) score, and distinct overlapping BC abnormalities, demonstrating moderate accuracy, sufficient calibration, and clinical benefits for prognostication.

In conclusion, sex-specific variations in BC profiles were observed among the patients with decompensated cirrhosis.

Figures