Circumstances such as age,1–5 gender,5,6 diabetes mellitus (DM)1,5,7 or kidney failure8,9 have an impact on patients’ body composition. In fact, older patients generally have less lean tissue than younger patients, women have less lean tissue than men, patients with DM usually have more fat tissue than non-diabetics, and patients on dialysis have less lean and fatty tissue than healthy people, this is likely the result of being in a hypercatabolic state10 with insufficient intake/absorption of nutrients.11–13

Therefore, given that a higher presence of lean tissues has been correlated with better survival14,15 in patients on haemodialysis (HD), reference values16 adapted to the HD population should be available. These values must take into account characteristics such as age, gender, or DM, in order to get an adequate assessment of body composition parameters such as the lean tissue index (LTI) or the fat tissue index (FTI).

The use of body composition monitors has been extended in the HD units because of its easy of use and because they have been validated in the healthy and HD populations.16–20 Body Composition Monitor (BCM) (Fresenius Medical Care, Bad Homburg, Germany) based on bioimpedance spectroscopy (BIS) will be used in the present study.

The aim is, firstly, to describe the body composition of 6395 patients and to calculate the 10th and 90th percentiles of LTI/FTI in 4 groups of patient (diabetic men, non-diabetic men, diabetic women, non-diabetic women) relative to age, so percentiles will serve as a reference to assess patient's lean and fat tissue content. Secondly, to evaluate the mortality of the population and determine if having a LTI below the 10th percentile in their corresponding group carries a higher risk of death in HD; and thus confirm that LTI/FTI reference values provided are valid.

The baseline characteristics of the cohort are shown in Table 1. The causes of kidney disease were: undetermined (31.5%), diabetic nephropathy (21.8%), glomerulonephritis (15.3%), nephroangiosclerosis (12.2%), tubulointerstitial nephropathy (9.5%), polycystic kidney disease (7.2%), and other (2.6%). As for vascular access: 31.8% of patients received dialysis through a catheter and 68.2% through an autologous or heterologous arteriovenous fistula.

Patients were grouped by gender and presence of DM. The 25th, 50th, and 75th percentiles of LTI and FTI were calculated for each of the four groups (Fig. 1).

Fig. 1.

The 25th, 50th, and 75th percentiles of fat tissue index (FTI) and lean tissue index (LTI) for the 4 groups.

(0.22MB).

The 10th and 90th percentiles were also calculated for each age decil in the four patient groups. Fig. 2 shows the patient distribution based on their LTI: those with an LTI below the 10th percentile or over the 90th percentile (black points) and those with an LTI between the 10th and 90th percentiles (grey points).

Fig. 2.

Patient distribution based on lean tissue index (LTI). Black points are those with an LTI below the 10th percentile or above the 90th percentile, grey points are those with an LTI between the 10th and 90th percentiles.

(0.68MB).

A total of 707 patients died during follow-up (11.1%), the majority of them from a cardiovascular cause (41.6%). Of the surviving patients, 79.1% continued HD in FMC centres, 10.8% transferred to non-FMC centres, 7.8% received transplants, 1.9% stopped renal replacement therapy, and 0.4% recovered kidney function.

The result from the survival analysis using multivariate logistic regression is shown in Table 2. It includes the factor “LTI below the 10th percentile of their group and age” and others such as: age, gender, vintage, Charlson index, vascular access, overhydration (ROH greater than 15%), and albumin.

The work provides LTI and FTI reference values (10th and 90th percentiles) adjusted for age, gender, and DM for the HD population, coming from the HD population. The descriptive analysis confirms the effect of some circumstances on the body composition of HD patients, such as age,1–4 gender6 and DM1,7 and, therefore, the need to assess patients’ body composition taking these circumstances into account.

The sample size, representing a substantial proportion of HD patients in Spain, is one of the work's strengths. This enabled us to obtain LTI and FTI reference ranges for HD patients coming from subjects with their same characteristics. To our knowledge to date, Until now, the reference values for the body composition parameters obtained using BIS came from healthy patients.6

Furthermore, it was observed that the presence of less lean tissue, defined a “LTI below the 10th percentile of their group” carries a higher risk of death in HD (OR 1.57), which also demonstrates that the reference values provided by the work are valid, illustrative, and adequate.

The study shows that the association between low LTI and mortality remains statistically significant, even after adjusting for other risk factors such as time in HD,23 vascular access,24 Charlson index,1,6,7 relative overhydration greater than 15%7,25–29 and albumin.23,30–33

Previous studies have examined the correlation between low lean tissue (expressed as LTI below the 10th percentile in the present study) and worse survival in HD34. Among the possible causes we note that (1) a lower lean mass may be a sign of malnutrition35,36 or a sign of inflammation;36 (2) that there may be uraemic toxin deposits in lean tissue, and therefore, lesser lean tissue may indicate a higher concentration of uraemic toxins in blood;36,37 (3) that the increase in lean tissue (and muscle mass) is associated with physical activity and this improves vascular rigidity and insulin resistance;1,36 and, lastly, (4) the known correlation between muscle mass and mortality caused by infections in HD patients, since chronic kidney disease patients have an acquired immunodeficiency that may become exacerbated in patients with little muscle mass by sharing risk factors such as hypoalbuminemia and protein-energy wasting (PEW).2

According to the possible explanations, we must add that patients with LTI below the 10th percentile present higher levels of CRP, but their dialysis doses are those recommended by the guidelines.38–41 Thus, the body composition of HD patients, and especially those with LTI below the 10th percentile, could reflect an underlying disease (inflammatory processes or malnutrition) or could perhaps predispose patients to risky situations, such as infections.

The results of the study suggest that the physicians should establish some diagnostic and therapeutic measures in HD patients with low LTI.42 Firstly, at attempt to identify the reason of low lean tissue and treat the underlying cause; secondly, to determine if nutritional intervention such as oral or intravenous supplement may be useful in this type of patient; and thirdly, to improve the dialysis dose if the cause is attributed to underdialysis.

This study shows the 10th and 90th percentile values adjusted by age, gender, and DM in a significant number of HD patients in Spain. These LTI/FTI values may be used by physicians as reference values to assess the nutritional status of HD patients. In fact, using these values as a guide, it was shown that the presence of a “LTI below the 10th percentile” for their group is correlated with a higher risk of death, even after adjusting for other factors such as overhydration.

One of the limitations of this study is that the patients’ race was not taken into account since the vast majority of patients were Caucasioan1; but reference values are needed for other races. Other limitations are that neither the interventions (such as nutritional support or physical activity) nor specific circumstances (such as hospitalizations), and their effects on body composition or on morbidity-mortality of the patients were recorded. This could be taken into account in future works. Similarly, as more patients are incorporated to total sample, also vintage can be considered a grouping factor.

In conclusion, new diagnostic tools such as BCM allow us a simultaneous assessment of hydration and nutrition situation of HD patients. Monitoring the body composition of patients and its assessing using adequate reference values may be useful to identify situations of risk that can be prevented or treated.

Now it remains to be shown in prospective studies if nutritional intervention with an increase in lean tissue may result in improved survival.

All the authors are employees of Fresenius Medical Care.