**2.4 Bioimpedance and body composition monitoring**

Bioimpedance monitors can provide quick, cheap, non-invasive bedside measurements of fluid status with good reproducibility. However, until recently the use of bioimpedance has been restricted to a relatively small number of centres with both a clinical and an academic interest in the technology.

In 'single frequency' bioimpedance monitors, a tiny 50 kHz alternating current is passed between a pair of electrodes, usually placed on the hand and foot. Sensing electrodes, placed just inside the current carrying electrodes, measure 'resistance' and 'reactance' to the passage of the current. Resistance and reactance combine to give the overall 'impedance'. An increase in body water makes it easier for current to pass through the body so that resistance decreases. Reactance, which is due to the capacitance of cell membranes, decreases as the number and/or integrity of the membranes decreases.

Single frequency bioimpedance monitors are widely used in health clubs as they can give an estimate of body fat and muscle mass. The technology is also built into bathroom scales where the current is passed between the feet. In dialysis patients, the equations used to derive body composition from the impedance at 50 kHz are unreliable if the patient has an abnormal fluid status.

Prof Antonio Piccoli and co-workers recognised this and developed bioimpedance vector analysis (BIVA) which simply looks at the hydration of the body tissues between electrodes placed on the hand and foot. The resistance and reactance measurements are normalised to height and interpreted using gender-specific nomograms derived from large studies of normal subjects and of populations with altered body composition (Piccoli 1994, 1995). As shown in figure 2, fluid overload is associated with movement of the vector downward and to the left whilst dehydration moves the vector up and to the right.

Bioimpedance monitors can provide quick, cheap, non-invasive bedside measurements of fluid status with good reproducibility. However, until recently the use of bioimpedance has been restricted to a relatively small number of centres with both a clinical and an academic

In 'single frequency' bioimpedance monitors, a tiny 50 kHz alternating current is passed between a pair of electrodes, usually placed on the hand and foot. Sensing electrodes, placed just inside the current carrying electrodes, measure 'resistance' and 'reactance' to the passage of the current. Resistance and reactance combine to give the overall 'impedance'. An increase in body water makes it easier for current to pass through the body so that resistance decreases. Reactance, which is due to the capacitance of cell membranes, decreases as the

Single frequency bioimpedance monitors are widely used in health clubs as they can give an estimate of body fat and muscle mass. The technology is also built into bathroom scales where the current is passed between the feet. In dialysis patients, the equations used to derive body composition from the impedance at 50 kHz are unreliable if the patient has an

Prof Antonio Piccoli and co-workers recognised this and developed bioimpedance vector analysis (BIVA) which simply looks at the hydration of the body tissues between electrodes placed on the hand and foot. The resistance and reactance measurements are normalised to height and interpreted using gender-specific nomograms derived from large studies of normal subjects and of populations with altered body composition (Piccoli 1994, 1995). As shown in figure 2, fluid overload is associated with movement of the vector downward and

Fig. 1. Variation in blood volume with extracellular volume

**2.4 Bioimpedance and body composition monitoring** 

number and/or integrity of the membranes decreases.

to the left whilst dehydration moves the vector up and to the right.

interest in the technology.

abnormal fluid status.

Fig. 2. Nomogram used for interpreting single frequency bioimpedance measurements. The outer ellipse encloses 95% of readings for normal subjects.

BIVA could provide practical information on changes in fluid status using very simple equipment that would cost little more than a set of bathroom scales if the market was larger. The reason BIVA cannot give an accurate indication of the patient's normally hydrated weight is clear when you examine Figure 2. The vector shown could be obtained from a slightly overweight subject with normal hydration, from a muscular subject with fluid overload or from an obese subject who is dehydrated. The confusion arises because adipose tissue contains very little intracellular water as fat cells are filled with triglycerides, but does have water in the extracellular space. So, like overhydration, an increase in body fat leads to an increase in the proportion of fluid in the extracellular space leading to a shorter vector.

In whole body bioimpedance spectroscopy (BIS), the electrodes are placed as for BIVA and resistance and reactance is measured over a range of frequencies. The results, together with the height, weight and gender of the subject, are used to compute the intracellular and extracellular water volumes (ECW and ICW). Fluid overload is associated with an increase in the proportion of water in the extracellular space but until recently, it was necessary for the user to decide what this proportion should be at normal hydration. There are a number of published methods for doing this (for example Lindley et al, 2005; Lopot et al, 2002) but they involve comparing dialysis patients with normal controls. As with BIVA, this makes it difficult to assess fluid status in patients with abnormal body composition.

Body composition monitoring (BCM) is the most recent commercially available development in bioimpedance monitoring. It uses the same electronic measurements as BIS to determine ECW and ICW but incorporates additional modelling (Moissl et al, 2006; Chamney et al, 2007) to take account of the amount of body fat the patient actually has,

Management of Fluid Status in

Haemodialysis Patients: The Roles of Technology and Dietary Advice 191

systematically decreasing their target weight could compromise their residual renal

When the BCM measurement is combined with the clinical indicators in Table 1, and knowledge of the patient's usual IDFG, it is possible to customise the target weight as described in 2.1. Introducing BCM has been shown to improve blood pressure control and reduce intradialytic adverse events (Machek, 2010). Ideally BCM should be carried out at least quarterly, though more frequent measurements will be needed for patients who are unwell (especially if admitted to hospital) or who are trying to gain or lose weight. As well as ensuring timely adjustments to the prescribed target weight, BCM gives valuable information on changes in body fat and lean tissue and provides an accurate urea

The patient's hydration status during the interdialytic period depends on both the weight achieved after dialysis and the fluid gained by the patient before the next session. Very high IDFG can make it impossible for the patient to remain close to normal hydration and to control pre-dialysis blood pressure. Another problem, particularly in elderly and malnourished patients, is the inability to tolerate the ultrafiltration rates required to remove a moderate volume of accumulated fluid. Whether the patient is gaining excessive volumes or failing to transfer fluid from the tissues into the circulation sufficiently rapidly, the

A typical haemodialysis patient in the UK accumulates about 2 litres of excess fluid in the intervals between sessions. When they attend for dialysis the machine is programmed to remove the excess fluid by ultrafiltration. Every litre removed in this way will carry with it about 137 mmol of sodium ions, though the actual amount will depend on the serum sodium level at the time the fluid was removed. As 137 mmol is the amount of sodium in 8 g of salt, the typical UK patient loses sodium equivalent to about 16 g of salt at each dialysis session. The body does have 'non-osmotic' sodium stores in tissues such as the skin and connective tissues (Titze, 2008) and it is possible that sodium can be recruited into or removed from these stores to buffer short term fluctuations in serum sodium. However, if the patient is assumed to be in steady state on the timescale of the interdialytic period, they must be making up for the sodium lost by taking in the equivalent of 16 g of salt between sessions. Sodium does come in other forms other than salt, such as sodium bicarbonate, but

If retained in the body, the salt taken in will cause 'osmometric' thirst. Osmometric thirst is triggered when increased osmolarity of the extracellular fluid causes osmoreceptor cells in the hypothalamus to shrink. Volumetric thirst, which occurs when the body loses both water and salt, is triggered when baroreceptors in the atria sense low cardiac return volume. Haemodialysis patients may experience volumetric thirst immediately after dialysis if they are dehydrated, but at other times their thirst is primarily osmometric. Fluid drunk in response to post-dialysis dehydration does not usually lead to increased IDFG as it simply delays osmometric thirst until the patient has consumed enough salt. An important exception to this occurs if a patient with good residual renal function is dehydrated, as they

will need to take in enough fluid to normalise their hydration before diuresis starts.

Osmometric thirst is part of the body's system for maintaining electrolyte balance. If our typical haemodialysis patient is anuric (unable to lose sodium via the kidneys), they will need to take in about one litre of water to dilute every 8g salt consumed to a normal

function, cause cramps and leave them feeling exhausted for hours after dialysis.

distribution volume for use in on-line measurements of dialysis adequacy.

**3.1 Salt and fluid balance: Osmometric thirst and the sodium 'set-point'** 

**3. Control of interdialytic fluid gain (IDFG)** 

solution is to try and reduce their IDFG.

it is usually combined with chloride.

rather than assuming they have the average amount for a person of their age and gender. Essentially, the BCM model assumes that the body is composed of normally hydrated lean tissue, normally hydrated adipose tissue and excess fluid (or missing fluid if the patient is dehydrated). For any combination of ECW, ICW and weight, there is only one matching combination of lean tissue, fat and excess/missing fluid.

With BCM, we can get an estimate of normal hydration and, for the first time, select a target weight that minimises the unwanted effects of dehydration as well as those of overhydration. Just as the introduction of BVM revealed a cohort of chronically overloaded patients, introducing BCM as part of the assessment of fluid status identified a group who were excessively dehydrated. In some of the cases identified in this way, an increase in target weight led to a reduction in interdialytic fluid gain as the patient's kidneys were able to produce more urine.

Fig. 3. Pre-dialysis systolic blood pressure vs hydration status

Users of BCM soon become aware that blood pressure can be misleading when evaluating fluid status. The scatter plot in Figure 3 shows the relationship between pre-dialysis systolic blood pressure and hydration status (the difference between measured and normal hydration in litres), for the first BCM measurement made in 474 haemodialysis patients under the care of Leeds Teaching Hospitals. A very similar plot was obtained in a crosssectional study of 639 PD patients (Van Biesen et al, 2010). The expected increase in blood pressure with fluid overload is present, but only as a trend for the population. For an individual patient, a single high or low blood pressure measurement is a rather poor predictor of fluid status, though a trend to higher or lower blood pressures in the same patient does provide important clinical information.

A significant number of patients who are normally hydrated, or even dehydrated, predialysis have high blood pressure. The traditional method for treating these patients by

rather than assuming they have the average amount for a person of their age and gender. Essentially, the BCM model assumes that the body is composed of normally hydrated lean tissue, normally hydrated adipose tissue and excess fluid (or missing fluid if the patient is dehydrated). For any combination of ECW, ICW and weight, there is only one matching

With BCM, we can get an estimate of normal hydration and, for the first time, select a target weight that minimises the unwanted effects of dehydration as well as those of overhydration. Just as the introduction of BVM revealed a cohort of chronically overloaded patients, introducing BCM as part of the assessment of fluid status identified a group who were excessively dehydrated. In some of the cases identified in this way, an increase in target weight led to a reduction in interdialytic fluid gain as the patient's kidneys were able

combination of lean tissue, fat and excess/missing fluid.

Fig. 3. Pre-dialysis systolic blood pressure vs hydration status

patient does provide important clinical information.

Users of BCM soon become aware that blood pressure can be misleading when evaluating fluid status. The scatter plot in Figure 3 shows the relationship between pre-dialysis systolic blood pressure and hydration status (the difference between measured and normal hydration in litres), for the first BCM measurement made in 474 haemodialysis patients under the care of Leeds Teaching Hospitals. A very similar plot was obtained in a crosssectional study of 639 PD patients (Van Biesen et al, 2010). The expected increase in blood pressure with fluid overload is present, but only as a trend for the population. For an individual patient, a single high or low blood pressure measurement is a rather poor predictor of fluid status, though a trend to higher or lower blood pressures in the same

A significant number of patients who are normally hydrated, or even dehydrated, predialysis have high blood pressure. The traditional method for treating these patients by

to produce more urine.

systematically decreasing their target weight could compromise their residual renal function, cause cramps and leave them feeling exhausted for hours after dialysis.

When the BCM measurement is combined with the clinical indicators in Table 1, and knowledge of the patient's usual IDFG, it is possible to customise the target weight as described in 2.1. Introducing BCM has been shown to improve blood pressure control and reduce intradialytic adverse events (Machek, 2010). Ideally BCM should be carried out at least quarterly, though more frequent measurements will be needed for patients who are unwell (especially if admitted to hospital) or who are trying to gain or lose weight. As well as ensuring timely adjustments to the prescribed target weight, BCM gives valuable information on changes in body fat and lean tissue and provides an accurate urea distribution volume for use in on-line measurements of dialysis adequacy.
