Accuracy and Utility of Estimating Lean Body Mass and Nutritional Status in Patients with Chronic Kidney Disease on Long-Term Hemodialysis Using Anthropometric Skinfold Thickness Measurements.
Lean body mass (LBM) assessment and subjective global assessment (SGA) are well-recognized methods to assess a patient's nutritional status. SGA involves a standardized clinical examination and routine assessment using well-established questionnaires (Riella, 2013). The dual-energy X-ray absorptiometry (DEXA) scan is a reliable method to assess body composition: fat mass, LBM, total body water, and bone density (Noori et al., 2011). It is regarded as a gold standard method in assessing LBM, and hence, nutritional status (Broers et al., 2015; Noori et al., 2011; Plank, 2005). However, DEXA is not easily available and is relatively costly. Simpler, less expensive, and more accessible techniques to assessing LBM are required. Anthropometric skinfold thickness measurement (ASFM) has been studied and is relatively accurate in assessing LBM (Beberashvili et al., 2013; Noori et al., 2011). LBM percentage (LBM%) can be easily calculated via the well-established Jackson and Pollock equation for AS FM (Jackson & Pollock, 2004; Jackson, Pollock, & Ward, 1980). It is simple, easily accessible, and relatively inexpensive. However, its accuracy is questioned in patients with fluctuating hydration status, such as those on chronic hemodialysis (Nafzger et al., 2015); therefore, it is hardly used.
We hypothesized that LBM% calculated from ASFMs would give a comparable estimate of LBM% as calculated by the whole body DEXA in patients on chronic hemodialysis. LBM% is a marker of nutritional status, and our aim was to assess the degree of correlation between the nutritional status as assessed by ASFM and SGA, in comparison with DEXA scan evaluation, in patients on chronic hemodialysis.
The aim of our study was to estimate the utility and accuracy of ASFMs in estimating LBM in patients on hemodialysis using DEXA scan as a gold standard in assessing LBM. Additionally, we aimed to assess the degree of correlation of LBM as assessed by ASFM and SGA.
Material and Methods
Study Design and Patients
Our research was a pilot prospective observational study conducted from August 2013 to November 2014. The Peninsula Health Ethics Committee approved this study. This study was conducted in accordance with the Declaration of Helsinki.
Inclusion criteria were those above age 18 years, and stable patients having chronic hemodialysis for more than three months and eligible to have DEXA scan as part of their regular review. Excluded patients were those under age 18 years, did not have the capacity to make informed consent, those having hemodialysis for less than three months, those hospitalized with an acute illness, limb amputees, and those not eligible to have a routine pharmaceutical benefit scheme (PBS)-funded DEXA scan.
Figure 1 Body Density, Body Fat Percentage, and Lean Body Mass Estimates Females: Body density = 1.0994921 - 0.0009929*X + 0.0000023*[X.sup.2] - 0.0001392*age (where X = sum of triceps, supra iliac and thigh skinfold measurements, and [X.sup.2] = square of the sums of triceps, supra iliac and thigh skinfold measurements) Males: Body density = 1.1093800 - 0.0008267*X + 0.0000016*[X.sup.2] - 0.0002574*age (where X = sum of chest, abdominal, thigh skinfold measurements, and [X.sup.2] = square of the sums of chest, abdominal, thigh skinfold measurements)
The whole-body DEXA scan was performed on non-dialysis days, midweek (Wednesday or Thursday), on all patients. The ASFM and SGA were performed on all patients immediately after their DEXA scan. The ASFM (index test) was then compared to the LBM established by the DEXA scan. Participants were also evaluated clinically regarding their fluid status at the same time of the ASFMs and SGA. Fluid status is known to impact the degree of fat-free mass (FFM) estimated by DEXA.
Anthropometric Skinfold Measurements and Subjective Global Assessment
ASFMs of the triceps, chest, abdominal, supra-iliac, and thighs were performed using a body calliper. The average measurement of the three was used for a specific site. The Jackson and Pollock equation was used to estimate the body density, body fat percentage (Fat%), and LBM (see Figure 1).
The Fat% was determined by the Siri equation, as below:
Fat% = [(495 / Body Density) - 450] * 100
Total LBM was estimated by subtracting total body fat from total body weight. The LBM% was then calculated.
LBM = Total Body Weight - Body Fat LBM% = (Total LBM/ Total Body Weight) * 100
SGA is a standardized questionnaire completed based on medical history and physical examination as shown below. Based on the assessment, the clinician rated each patient as follows: SGA Category A = well-nourished, SGA Category B = mildly malnourished or suspected of malnutrition, or SGA Category C = severely malnourished.
* Medical history: weight change; dietary intake; gastrointestinal symptoms; functional impairment.
* Physical examination: loss of subcutaneous fat; muscle wasting; edema; ascites at fat pads of the eye, triceps skinfold, biceps skinfold, temporal area, clavicular area, shoulder joint, scapula, ribs, interosseous muscle, quadriceps, and calf muscles.
ASFMs and SGA assessments were repeated in three to six months. ASFMs and the SGA were taken by three investigators assigned in this study to minimize inter-observer variability.
Body Compositions by DEXA (Reference Test)
All participants had their whole-body DEXA performed using one assigned DEXA scanner (Medilink MedixDR). Routine hip and spine, as well as an additional whole-body scan, were examined during the DEXA scan. Calculation by DEXA scan using the 3-compartment (3C) model is a gold standard in body composition measurements. Body compositions estimated and provided by DEXA were fat mass (FM), FFM, and bone mineral density (BMD).
We examined a plot of the difference against the average (Bland-Altman plot) (Altman & Bland, 1996) to assess the agreement between LBM measured by DEXA and ASFM, and the repeatability of LBM measured by ASFM at two time points. A scatter plot evenly distributed above and below the middle line (the line of no difference) indicates there is no systematic bias between the two measurements. We also estimated the mean difference and its associated 95% confidence interval (95% CI), the limits of agreement, and intra-class correlation (ICC) describing the relative extent to which the two ASFMs are related. An ICC value of 0.95 indicates that 5% of the variance is due to measurement error or the variance within the two time points. To assess the correlation between the ASFM and SGA, we fitted a linear regression model of LBM measured by ASFM as the outcome and SGA as the exposure of interest.
A total of 25 patients participated in our study (see Table 1). The mean age was 61.2 years (Standard Deviation [SD]=10.4 years), and 15 (60%) participants were men. The average weight and body mass index (BMI) were 87.63 kg (SD=20.3kg) and 29.69 kg/[m.sup.2] (SD=6.31 kg/[m.sup.2]), respectively. The average Fat% for our cohort was 25.45 [+ or -] 7.18% (22.02 [+ or -] 4.89% for men; 30.08 [+ or -] 7.56% for women).
Fifteen participants had a follow-up ASFM and SGA in 3 to 6 months from their initial measurement. Their mean age was 60.93 [+ or -] 12.7 years, and 10 (67%) participants were men. This group had a mean weight of 87.00 [+ or -] 17.88 kg and BMI average of 31.02 [+ or -] 5.05 kg/[m.sup.2].
Of the 25 patients at baseline, 21 were clinically euvolemic. Twelve of 15 patients who had a follow-up ASFM and SGA were clinically euvolemic.
Comparability of LBM% Measured by DEXA and ASFM
The average LBM% calculated by ASFM and DEXA at baseline was 70.29 [+ or -] 7.09% and 71.75 [+ or -] 6.51%, respectively. The mean difference of LBM% calculated by DEXA compared to ASFM was -1.46% (95% CI: -4.09 to 1.18). Thus, ASFM can give up to a 4.09% lower reading than DEXA or up to a 1.18% higher reading compared to DEXA. Limits of agreement (reference range for difference) range from -14.0 to 11.1 and an ICC of 0.71. Figure 2 shows a comparison of LBM% measured by DEXA and by ASFM using the Bland-Altman plot. The even distribution of the Bland-Altman plot in Figure 2 demonstrates that correlation of LBM% using ASFM compared to DEXA had good limits of agreement with close ICC between both methods in our study.
Comparability Between LBM% and Nutritional Status by ASFM and SGA
Twenty-four participants had SGA data available. The average LBM% measured by skinfold for those with SGA Category A was 68.0%. Table 2 shows results of a linear regression analysis. LBM% had a 2.46% difference (95% CI: -3.36, 8.28; p=0.389) for those with SGA Category B compared to SGA Category A. LBM% difference for SGA Category C compared to SGA Category A was 16.16% (95% CI: 7.61, 24.71; p=0.001). However, only two participants had an SGA Category C. Assessment of nutritional status by SGA of the severely malnourished group (SGA Category C) had a significant correlation with the nutritional status (LBM%) calculated by ASFM, but it was not found to be significant in those who were mildly malnourished or suspected of malnutrition (SGA Category B).
Repeatability of LBM by Skinfold Measured at Time 1 and Time 2
A comparison of LBM% measured by ASFM at two different times is shown in Figure 3 (baseline and repeated in three to six months' time). The mean difference between LBM measured at the two time points is 2.4% (95% CI: -4.9 to 0.2). Limits of agreement range from -11.3 to 6.6, and the ICC is 0.88.
A large proportion of patients on chronic hemodialysis are malnourished, with prevalence ranging from 30% to 50% (Riella, 2013). This is often termed as "protein energy wasting" (PEW) (Riella, 2013). The International Society of Renal Nutrition and Metabolism (ISRNM) panel proposed four main diagnostic criteria for detection of PEW: 1) biochemical markers, 2) body mass and composition, 3) muscle mass, and 4) dietary intake (Fouque et al., 2008). Deficiencies in three out of four of these categories are needed for a diagnosis of PEW (Fouque et al., 2008).
Sarcopenia is defined as loss of muscle mass and strength. PEW entails with it a substantial increase in morbidity and mortality in this patient population (Beberashvili et al., 2013; Kovesdy & Kalantar-Zadeh, 2012). This is often multi-factorial with reduced oral nutritional intake, metabolic acidosis, chronic inflammatory state, and increased catabolism from co-morbid conditions (Pellicano, Strauss, Polkinghorne, & Kerr, 2011).
The term malnutrition among individuals on maintenance hemodialysis has long been debated, and to date, has no clear definition. Not surprisingly, there is no pre-specified classification of target LBM% for patients on chronic hemodialysis. The recommended Fat% based on a general adult population is roughly 25% to 31% for females and 18% to 24% for males (Bays et al., 2016-2017). This was consistent with our cohort of patients. Therefore, we used the recommended body composition as a reference and guide to assessing nutritional status in our study cohort.
Our study aimed to identify an easier and accessible method of measuring and monitoring LBM% and the nutritional status of patients on chronic hemodialysis. This is to allow early identification of undernourished individuals, allowing earlier relevant referrals and intervention to avoid the associated morbidity and mortality with malnutrition. To date, studies have utilized other ways of calculating LBM, such as by mid-arm circumference (MAMC) and bio-impedance analysis (BIA) (Beberashvili et al., 2013). These methods are not easily available and are too costly to be implemented routinely. LBM% by ASFM is a simple and potentially cost-effective way of assessing LBM%.
We found that LBM% measured by ASFM was comparable to LBM% estimated by whole-body DEXA, with the mean difference being--1.46% (95% CI -4.09 to 1.18), and with acceptable limits of agreement of LBM% by ASFM and DEXA ranging from -14.0 to 11.1. A reliable intraclass correlation of LBM% between ASFM and DEXA was 0.71. This suggests the LBM% of our participants were reliably measured by ASFM comparing to DEXA, a gold standard of measurement of LBM%. It highlights the ability to use ASFM to monitor nutritional status of patients on chronic hemodialysis accurately.
The SGA is widely used and is a helpful tool to assess the nutritional status of patients on chronic hemodialysis (Riella, 2013). The LBM% calculated by ASFM and DEXA was substantially correlated in our study. Utilizing the SGA to assess nutritional status, those under SGA Category A (well-nourished) had a mean LBM% of 68% (comparable LBM% to those well-nourished in the general population). Based on LBM% as a nutritional marker, SGA was only useful for the detection of patients who were severely malnourished. SGA was not useful for detecting those who were in SGA Category B (mildly malnourished). However, because only two patients in our study were rated as severely undernourished based on SGA, this skews this statistical finding. Based on our study, SGA may be helpful to detect only those who are severely malnourished. However, this demonstrates that SGA lacks the ability to detect those who are mildly undernourished or suspected of malnutrition compared to ASFM. Therefore, LBM% by ASFM would detect under-nourished patients early and allow for earlier interventions.
To allow for follow-up assessment of nutritional status and LBM% on our cohort, we repeated the ASFM and SGA at 3 to 6 months from the initial assessment. With only 15 participants with SGA measured at times 1 and 2, and three possible categories for rating SGA, there were little data to make a valid inference about the reproducibility of SGA. LBM% by ASFM was consistent, with a mean difference in LBM% of -2.4% (95% CI: -4.9 to 0.2), and with limits of agreement ranging from -11.3 to 6.6 and significant intra-class correlation of 0.88. LBM% calculated by ASFM may be reliably and accurately repeated when performed appropriately. Our cohort of patients was stable patients on chronic maintenance hemodialysis, and we did not expect their nutritional state to differ much in the subsequent 3 to 6 months.
There were several limitations in our study. First, this was a pilot study with a small cohort of patients recruited. This would impact the power of statistical findings. Second, patients on chronic hemodialysis likely have other co-morbidities impacting nutritional status, morbidity, and mortality. These data were not included in our pilot study. Third, we used the recommended guidelines of LBM% for the general population because no estimates exist for patients on chronic maintenance hemodialysis. Future studies could be done to hopefully estimate the recommended LBM% for patients on chronic maintenance hemodialysis to use this variable as a tool in estimating nutritional status. Further, our dataset was too small to make valid inferences by sex. No intervention was initiated during this study when patients were detected as "malnourished" based on LBM% calculated by ASFM or DEXA. The pilot study aimed primarily to detect the utility of ASFM in determining LBM% as compared to DEXA scan.
Our study demonstrated that ASFM is a useful way of assessing LBM% and nutritional status. There is a strong correlation between LBM% as measured by ASFM and the DEXA xcan. Repeated ASFM may be more sensitive than SGA in detecting malnutrition. ASFM is simpler and more accessible compared to other gold standard methods of calculating LBM%. This method has the potential to be used clinically to estimate and monitor the nutritional status of individuals on chronic hemodialysis. This allows for early identification and management of malnourished patients on chronic hemodialysis.
Altman, D.G., & Bland, J.M. (1996). Detecting skewness from summary information. British Medical Journal, 3/3(7066), 1200.
Bays, H.E., Seger, J.C., Primack, C., McCarthy, W., Long, J., Schmidt, S.L., ... Westman, E.C. (2016-2017). Obesity algorithm, presented by the Obesity Medicine Association. Retrieved from http://www.amga.org/wcm/ PI/Collabs/OCMC/ResourceGuide/ OMA_Obesity-Algorithm.pdf
Beberashvili, I., Azar, A., Sinuani, I., Kadoshi, H., Shapiro, G., Feldman, L., ... Weissgarten, J. (2013). Comparison analysis of nutritional scores for serial monitoring of nutritional status in hemodialysis patients. Clinical Journal of the American Society of Nephrology, 8(3), 443-451. doi:10.2215/ cjn.04980512
Broers, NJ., Martens, RJ., Cornelis, T., Diederen, N.M., Wabel, P., van der Sande, F.M., . Kooman, J.P. (2015). Body composition in dialysis patients: A functional assessment of bioimpedance using different prediction models. Journal of Renal Nutrition, 25(2), 121-128. doi:10.1053/ j.jrn.2014.08.007
Fouque, D., Kalantar-Zadeh, K., Kopple, J., Cano, N., Chauveau, P., Cuppari, L., ... Wanner, C. (2008). A proposed nomenclature and diagnostic criteria for protein-energy wasting in acute and chronic kidney disease. Kidney International, 73(4), 391-398. doi:10.1038/sj.ki.5002585
Jackson, A.S., & Pollock, M.L. (2004). Generalized equations for predicting body density of men. 1978. British Journal of Nutrition, 9/(1), 161-168.
Jackson, A.S., Pollock, M.L., & Ward, A. (1980). Generalized equations for predicting body density of women. Medicine & Science in Sports & Exercise, 12(3), 175-181.
Kovesdy, C.P., & Kalantar-Zadeh, K. (2012). Accuracy and limitations of the diagnosis of malnutrition in dialysis patients. Seminars in Dialysis, 25(4), 423-427. doi:10.1111/j.1525139X.2012.01097.X
Nafzger, S., Fleury, L.A., Uehlinger, D.E., Pluss, P., Scura, N., & Kurmann, S. (2015). Detection of malnutrition in patients undergoing maintenance haemodialysis: A quantitative data analysis on 12 parameters. Journal of Renal Care, 4/(3), 168-176. doi:10. 1111/jorc.12125
Noori, N., Kovesdy, C.P., Bross, R., Lee, M., Oreopoulos, A., Benner, D., ... Kalantar-Zadeh, K. (2011). Novel equations to estimate lean body mass in maintenance hemodialysis patients. American Journal of Kidney Disease, 57(1), 130-139. doi:10.1053/ j.ajkd.2010.10.003
Pellicano, R., Strauss, B.J., Polkinghorne, K.R., & Kerr, P.G. (2011). Longitudinal body composition changes due to dialysis. Clinical Journal of the American Society of Nephrology, 6(7), 1668-1675. doi:10.2215/cjn.06790810
Plank, L.D. (2005). Dual-energy X-ray absorptiometry and body composition. Current Opinion in Clinical Nutrition & Metabolic Care, 8(3), 305-309.
Qureshi, A.R., Alvestrand, A., Divino-Filho, J.C., Gutierrez, A., Heimburger, O., Lindholm, B., & Bergstrom, J. (2002). Inflammation, malnutrition, and cardiac disease as predictors of mortality in hemodialysis patients. Journal of the American Society of Nephrology, 13(Suppl. 1), S28-S36.
Riella, M.C. (2013). Nutritional evaluation of patients receiving dialysis for the management of protein-energy wasting: What is old and what is new? Journal of Renal Nutrition, 23(3), 195-198. doi:10.1053/j.jrn.2013.01.023
Khai Gene Leong
Jia Lian Chee
Leong, K.G., Chee, J.L., Karahalios, A., Skelley, A., & Wong, K. (2018). Accuracy and utility of estimating lean body mass and nutritional status in patients with chronic kidney disease on long-term hemodialysis using anthropometric skinfold thickness measurements. Nephrology Nursing Journal, 45(1), 35-40.
Khai Gene Leong, MBBS, Bmedsc, FRACP, is Consultant Nephrologist, Department of Nephrology, Monash Medical Centre, Clayton, Victoria, Australia.
Jia Lian Chee, MBBS, FRACP, is a Consultant Nephrologist, Department of Nephrology, Peninsula Health, Frankston, Victoria, Australia.
Amalia Karahalios, PhD, MPH, is a Research Fellow in Biostatistics, Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, University of Melbourne, and Office for Research, Western Centre for Health Research and Education, Sunshine, Victoria, Australia.
Annabelle Skelley, MBBS, MPHTM, is a Radiology Advanced Trainee, Department of Radiology, The Alfred Hospital, Melbourne, Victoria, Australia.
Kim Wong, FRACP, is a Consultant Nephrologist, Department of Medicine, Peninsula Health, Frankston, Victoria, Australia.
Statement of Disclosure: The authors reported no actual or potential conflict of interest in relation to this continuing nursing education activity.
Note: The Learning Outcome, additional statements of disclosure, and instructions for CNE evaluation can be found on page 41.
Caption: Figure 2 Lean Body Mass Measured by Skinfold and DEXA
Caption: Figure 3 Lean Body Mass Measured by Skinfold at Two Time Points
Table 1 Baseline Demographic Details of Participants Age Sex DEXA DEXA ASFM Overall BMI (Years) LBM% Fat% LBM% SGA (kg/[m.sup.2]) 61 M 72.00 24.97 70.96 B 29.07 49 M 67.40 30.38 73.79 B 28.41 71 F 61.65 36.44 64.17 A 33.20 61 F 59.71 38.69 64.83 A 33.91 48 F 79.60 16.99 87.94 C 18.78 60 M 79.64 16.34 67.13 A 26.54 71 F 64.89 32.27 73.83 A 21.94 78 F 70.06 28.00 61.22 A 31.11 56 M 73.24 24.19 75.58 A 35.44 63 M 72.92 23.88 68.28 A 25.65 52 F 77.91 18.34 80.44 C 20.43 42 F 60.35 37.95 56.61 A 42.77 82 M 78.67 17.51 70.84 B 26.83 48 F 64.29 33.86 60.52 A 39.45 63 M 70.00 26.95 72.89 B 33.95 62 M 73.09 23.96 65.11 A 36.07 81 M 75.11 20.94 80.07 -- 24.68 49 M 83.54 12.63 75.84 A 28.37 57 M 80.74 15.35 70.46 A 29.06 58 M 76.07 20.73 66.77 A 28.93 58 M 72.73 24.55 63.95 B 42.75 61 M 74.92 21.00 76.64 A 25.98 62 F 66.92 30.36 65.41 A 26.99 64 M 69.89 26.93 74.04 A 29.06 72 F 68.47 27.94 70.03 A 23.16 Notes: M = male, F = female, DEXA LBM% = lean body mass percentage by DEXA, DEXA Fat% = fat mass percentage by DEXA, ASFM LBM% = lean body mass percentage by anthropometric skinfold measurements, Overall SGA = overall subjective global assessment, BMI = body mass index. Table 2 Results of Linear Regression Analysis of LBM Measured by Skinfold and SGA Category SGA Number of Estimated b p-Value Category Participants Coefficient (95% CI) A 17 Ref -- B 5 2.46 (-3.36, 8.28) 0.389 C 2 16.16 (7.61, 24.72) 0.001
|Printer friendly Cite/link Email Feedback|
|Author:||Leong, Khai Gene; Chee, Jia Lian; Karahalios, Amalia; Skelley, Annabelle; Wong, Kim|
|Publication:||Nephrology Nursing Journal|
|Date:||Jan 1, 2018|
|Previous Article:||Comfort and Fluid Retention in Adult Patients Receiving Hemodialysis.|
|Next Article:||Factors Influencing the Introduction of a Process of Advance Care Planning in Outpatient Hemodialysis Facilities.|