Serum creatinine improves body mass index survival prediction in hemodialysis patients: a 1-year prospective cohort analysis from the ARNOS study

J Ren Nutr. 2011 Sep;21(5):369-75. doi: 10.1053/j.jrn.2010.08.005. Epub 2011 Jan 15.


Objective: This study sought to better characterize the relationships between body mass index (BMI) and lean body mass (LBM) as assessed by serum creatinine (SCr) and mortality.

Design: The data were collected from a prospective prevalent cohort in maintenance hemodialysis patients.

Setting: The study was carried out in 25 dialysis units in Rhônes Alpes area (France and Switzerland).

Patient: A total of 1,205 patients were followed up for 1-year, starting July 1, 2005.

Outcome: Mortality as well as clinical and biological routine parameters were recorded. Kaplan-Meier, Cox model, Log rank test were used for the statistical analysis.

Results: We found that SCr was a strong predictor of mortality (P < .001), whereas BMI was not. Additionally, higher BMI lost its protective effect when it was associated with low SCr. Survival was strongly reduced in patients having a predialysis SCr <717 μmol/L in patients with a BMI >23 (P < .001).

Conclusion: BMI should not be used by itself but in conjunction with SCr as a surrogate of LBM to improve its morbid-mortality predictive power. LBM should also be taken into account in further survival studies carried out in hemodialysis patients.

MeSH terms

  • Aged
  • Aged, 80 and over
  • Body Composition
  • Body Mass Index*
  • Creatinine / blood*
  • Data Collection
  • Female
  • France
  • Humans
  • Kaplan-Meier Estimate
  • Kidney Failure, Chronic / physiopathology
  • Kidney Failure, Chronic / therapy
  • Male
  • Middle Aged
  • Proportional Hazards Models
  • Prospective Studies
  • Renal Dialysis / mortality*
  • Serum Albumin / analysis
  • Switzerland


  • Serum Albumin
  • Creatinine