15-04-2024 | Ascites | Original Article
Longitudinal changes in bone mineral density may be associated with long-term survival in patients with cirrhosis: A proof of concept of study
Published in: Indian Journal of Gastroenterology
Login to get accessAbstract
Introduction
The impact of longitudinal changes in different body components measured via body composition analysis (BCA) on liver-related outcomes in patients with cirrhosis is poorly understood. We evaluated the prognostic relevance of longitudinal changes in body composition over one year in patients with cirrhosis.
Methods
This was a follow-up study of a randomized controlled trial evaluating changes in bone density measured via dual energy X-ray absorptiometry (DEXA) upon vitamin D supplementation. Patients with available anthropometric indices, fat mass (FM), fat-free mass (FFM), bone-density at lumbar spine (LD) and left femur-neck (FD) (assessed by T score) at two time points one year apart were assessed for outcomes. The prognostic relevance of change in parameters such as ΔFM, ΔFFM, ΔLD and ΔFD over one year was assessed and compared with baseline model for end-stage liver disease (MELD) score.
Results
Patients with cirrhosis (n=112) (mean age 41.8±12 years, 58.5% males) were followed up for median duration of 5.7 years interquartile range [IQR 3.5–5.7], with five-year survival rate of 77%. On serial BCA, ΔLD (p=0.029) and ΔFD (p=0.003) emerged as significant predictors of survival, whereas ΔFM (p=0.479), ΔFFM (p=0.245) and ΔBMI (p=0.949) were not. The area under curve of ΔLD and MELD score for predicting survival was 0.636 (0.5–0.773) and 0.664 (0.555–0.773), respectively. ΔFD<0.1 over one year had sensitivity and specificity of 70.4% and 56.5% to predict poor survival. The combination of ΔFD, MELD and ascites predicted five-year survival with an optimism-corrected c-statistic of 0.785.
Conclusion
Among body composition parameters, changes in bone mineral density correlate best with survival and have prognostic relevance similar to that of ascites and MELD score.
Graphical Abstract
×
![]()