Total body protein status assessed by different estimates of fat-free mass in adult peritoneal dialysis patients

D. J. Borovnicar, K. C. Wong, P. G. Kerr, D. B. Stroud, D. W. Xiong, B. J G Strauss, R. C. Atkins

Research output: Contribution to journalArticleResearchpeer-review

27 Citations (Scopus)


Objective: To assess the usefulness of fat-free mass (FFM) as an index of total body protein (TBPr) status in continuous ambulatory peritoneal dialysis (CAPD) patients. Design: TBPr was measured by in vivo neutron activation analysis (IVNAA) and expressed as a standardised protein index (PI). FFM was estimated by dual energy X-ray absorptiometry (DXA), whole body counting of total body potassium (TBK), and creatinine kinetics (CK), and expressed as a standardised FFM index (FFMI). FFM was also determined by a criterion method based on a four compartment model (4CM) which is defined as the sum of total body water determined by D2O dilution, TBPr determined by IVNAA, bone mineral determined by DXA, and glycogen estimated to be 4.4% of TBPr. Each patient was measured within a four hour period by all methods. Setting: Body Composition Laboratory, Monash Medical Centre. Subjects: Six male and twelve female CAPD patients (33-77 years). Results: FFMI assessed by DXA and by TBK agreed with measurements of PI on identifying the mean TBPr status of the CAPD group as significantly below a comparable normal reference population (mean Z score: PI = -1.01 (P < 0.05); FFMI by DXA = -0.50 (P < 0.05); FFMI by TBK = - 1.24 (P < 0.05)). In contrast, FFMI assessed by CK did not reveal a significantly reduced TBPr status (mean Z score: -0.70 (NS)). Furthermore, significant linear correlations were noted between PI and FFMI estimated by DXA and by TBK (r = 0.57 (P < 0.05) vs r = 0.69 (P < 0.005)) however no significant correlation was observed between PI and FFMI estimated by CK (r = 0.36 (NS)). Moderate variation in FFM hydration did not compromise the ability of DXA, TBK or CK to differentiate between protein depleted, normal and enriched patients. Comparison of FFM estimates between the criterion method and either DXA, TBK or CK revealed no significant bias (+ 1.8 kg vs -2.0 kg vs + 0.8 kg) and respective SEE values of 3.8 kg (8.3%), 5.9kg (14.3%) and 9.6 kg (21.7%). Conclusion: The findings of this study indicate that FFM estimated by either DXA or the whole body counting of TBK is a useful index of TBPr status in CAPD patients. However, FFM assessed by CK does not appear to be an appropriate index of TBPr status in CAPD patients.

Original languageEnglish
Pages (from-to)607-616
Number of pages10
JournalEuropean Journal of Clinical Nutrition
Issue number9
Publication statusPublished - 1 Sep 1996


  • CAPD
  • Creatinine kinetics
  • Dual energy X-ray absorptiometry
  • Fat-free mass
  • Total body potassium
  • Total body protein

Cite this