Developing and Validating New Models to Estimate Lean Body Mass in Patients with CKD

Protein-energy wasting (PEW), a condition associated with adverse outcomes due to cardiac comorbidity and inflammation, is a common complication in patients with CKD. Patients undergoing dialysis experience PEW most often, but PEW is also seen in non–dialysis-dependent CKD patients. As renal function declines, protein catabolism can be gradually aggravated via various complex mechanisms and may become apparent in nutritional indices such as somatic and visceral protein storage.

Reduced lean body mass (LBM) is a key index for somatic protein deficit and is a predictor of high mortality in CKD patients not on dialysis. Accurately measuring LBM during routine care can identify PEW, allowing for initiation of intervention, particularly in patients with CKD stage 3 to 5.

However, according to Xue Tian, MD, and colleagues, measurements for LBW are not routinely conducted due to the lack of accurate and simple measurement techniques. The gold-standard (tracer dilution) and reference (dual energy x-ray absorptiometry [DEXA]) techniques are laborious, invasive, and not suitable for routine care. Currently, there are no LBM estimation methods that are simple, practical, and reliable.

Recent studies in patients receiving hemodialysis have demonstrated similarity between LBM as measured by DEXA and LBM estimated using hand-grip strength and mid-arm muscle circumference measurements. Dr. Tian et al. developed and validated two new equations for estimating LBM in patients on peritoneal dialysis based on hand-grip strength and mid-arm muscle circumference. The equations have good precision and accuracy and small bias; they also demonstrate improved performance over creatinine kinetics and anthropometry methods.

Equations for estimating LBM in patients with non–dialysis-dependent CKD are also needed. The researchers conducted a study to develop equations for that patient population (patients with CKD stage 3 to 5), based on mid-arm muscle circumference and hand-grip strength. The equations were validated using comparisons to LBM measured by DEXA as the reference method. Study results were reported in the Journal of Renal Nutrition [2018;28(3):156-164].

The researchers recruited 300 patients with stage 3 to 5 non–dialysis-dependent CKD who were divided into two groups: 150 in the development group and 150 in the validation group. There were no differences between the two groups in demographic data, distribution of CKD, stage, and comorbidity (P<.05). With the exception of mean hemoglobin, laboratory measurements were also similar between groups (P<.05) (mean hemoglobin was higher in the validation group).

Mid-arm muscle circumference was significantly higher in the validation group compared with patients in the development group (P<.001). However, LBM, determined using DEXA, hand-grip strength for two sides, and dietary protein and energy intake, was comparable between the groups. There were no significant differences in volume status, including extracellular water, total body water, or the ratio of extracellular to total body water, between the groups.

Using the development dataset, the researchers constructed two equations for LBM estimation: LBM-H, based on hand-grip strength (defined as the dominant hand-grip strength), and LBM-M, based on mid-arm muscle circumference. Potential variables for the regression equations from age, sex, height, and weight were selected via performance of stepwise procedures.  In Spearman correlation analyses, LBM-DEXA was correlated significantly with sex, height, weight, hand-grip strength (r=0.72), and mid-muscle arm circumference (r=0.66; all P<.001). There were no correlations between LBM-DEXA and serum albumin or daily protein intake. Two multiple regression equations were established using hand-grip strength or mid-arm muscle circumference in combination with the selected demographic variables of sex, height, and weight. The R-square values were 0.900 for LBM-H and 0.894 for LBM-M.

The two formulas were applied to the validation group and compared with measurements made utilizing LBM-DEXA. Estimated LBM values using LBM-H and LBM-M were numerically close to those measured with DEXA (44.6 kg) and significantly correlated with them (P<.001 for both). There was significant correlation with hand-grip strength and mid-arm muscle circumference with values estimated using the equations and values measured using LBM-DEXA (P<.001) for all. There was no significant correlation with serum creatinine, daily protein, and energy intake (as observed in the development dataset). LBM-H and LBM-M, but not LBM-DEXA, were also significantly correlated with serum albumin.

In further assessment of the performances of the new equations, still using LBM-DEXA as the reference method, the LBM was slightly overestimated with both the LBM-H and LBM-M formula. The analyses were repeated in mutually exclusive strata to compare the performance of the equations in varying ranges of LBM. Analyses were conducted in strata of higher and lower than the LBM-DEXA median of 44.6 kg; at CKD stage 3 and CKD stages 4 and 5; and higher and lower than the ratio of extracellular to total body water. The observed biases were consistent across all the groups (P<.001-.05).

Generally, the interquartile range differences with both equations were small (3.92-5.32 kg) and were independent of the values of LBM-DEXA, the ratio of extracellular to total body water, and CKD stage, indicating the preciseness of the two new equations.

Limitations to the study cited by the authors included developing the prediction model in single-center non-dialyzed CKD patients, possibly limiting the generalizability of the findings, and including only clinically stable patients in the cohort.

“In summary, two new models for predicting LBM using hand-grip strength and mid-arm muscle circumference were developed and then validated in a relatively large sample of nondialyzed CKD patients. Results of the validation indicated that the equations can provide reliable and accurate estimates of LBM in nondialyzed CKD patients in clinical practice. Further studies are needed to validate the models in a larger study population, and longitudinal studies are required to evaluate the suitability of the formulae for detecting changes in LBM in response to the intervention,” the researchers concluded.

Takeaway Points

  1. Researchers developed and tested two equations that can be used in daily clinical practice for the estimation of lean body mass (LBM) in patients with non–dialysis dependent CKD.
  2. One model was based on hand-grip strength and the other was based on mid-arm muscle circumference; validation was conducted using dual-energy x-ray absorptionmetry as the reference tool.
  3. Both models provided accurate estimates of LBM in nondialyzed CKD patients