In this protocol, we explain how to obtain and interpret phase angle values and bioelectrical impedance vectorial analysis (BIVA) Z-score obtained by bioelectrical impedance in patients with acute heart failure admitted to the Emergency Department and their clinical applicability as a predictive marker for the prognosis of a 90-day event.
Acute heart failure is characterized by neurohormonal activation, which leads to sodium and water retention and causes alterations in body composition, such as increased body fluid congestion or systemic congestion. This condition is one of the most common reasons for hospital admission and has been associated with poor outcomes. The phase angle indirectly measures intracellular status, cellular integrity, vitality, and the distribution of spaces between intracellular and extracellular body water. This parameter has been found to be a predictor of health status and an indicator of survival and other clinical outcomes. In addition, phase angle values of <4.8° upon admission were associated with higher mortality in patients with acute heart failure. However, low phase angle values may be due to alterations-such as the shifting of fluids from an intracellular body water (ICW) compartment to an ECW (extracellular body water) compartment and a concurrent decrease in body-cell mass (which can reflect malnutrition)-that are present in heart failure. Thus, a low phase angle may be due to overhydration and/or malnutrition. BIVA provides additional information about the body-cell mass and congestion status with a graphical vector (R-Xc graph). In addition, a BIVA Z-score analysis (the number of standard deviations from the mean value of the reference group) that has the same pattern as that of the ellipses for the percentiles on the original R-Xc graph can be used to detect changes in soft-tissue mass or tissue hydration and can help researchers compare changes in different study populations. This protocol explains how to obtain and interpret phase angle values and BIVA Z-score analyses, their clinical applicability, and their usefulness as a predictive marker for the prognosis of a 90-day event in patients admitted to an emergency department with acute heart failure.
Acute heart failure (AHF) results from the rapid onset of signs, symptoms, and exacerbation of derivates of HF and a combination of clinical, hemodynamic, and neurohormonal abnormalities, including systemic inflammatory activation, which leads to sodium and water retention1. This long-term accumulation causes the interstitial glycosaminoglycan (GAG) networks to become dysfunctional, resulting in reduced buffering capacity and changing the form and function of the GAG networks1,2. This contributes to alterations in body composition due to the shifting of fluids from intracellular to extracellular space3, thus inducing an increase in body fluids and leading to congestion, which is the most common cause of hospitalization with HF. It is principally fluid overload, compartmental fluid redistribution, or a combination of both mechanisms that require immediate medical attention4,5. This condition is one of the main predictors of a poor prognosis6,7.
Considering that AHF is the most common cause of hospital admissions in patients older than 65 years of age8, around 90% of those who are admitted to an emergency department present fluid overload6, and approximately 50% of these patients are discharged with persistent symptoms of dyspnea and fatigue, and/or minimal or no weight loss9. In-hospital mortality rates range from 4% to 8% after discharge; there is an increase from 8% to 15% at three months, and for re-hospitalization, the rates range from 30% to 38% at 3 months10. Therefore, the quick and accurate evaluation of congestion in real-time and acute settings, such as an emergency department, is crucial for therapeutic management11 and determining disease prognosis, morbidity, and mortality6.
Bioelectrical impedance analysis (BIA) has been suggested for estimating body composition for being safe, noninvasive and portable techinque12. To estimate a whole-body impedance, BIA uses a phase-sensitive impedance analyzer that introduces a constant alternating current through tetrapolar surface electrodes placed on the hands and feet12. This method combines the resistance (R), reactance (Xc), and phase angle (PhA)13, where R is the opposition to the flow of the alternating current through the intracellular and extracellular ionic solution. Xc is the delay in the conduction (dielectric components) or compliance of the tissue interfaces, cell membranes, and organelles with the passage of the administered current12. The PhA reflects the relationship between R and Xc. It is derived from the electrical properties of the tissue; it is expressed as the lag between the voltage and current at the cell membrane and tissue interfaces and is measured with phase-sensitive devices14,15,16,17.
The PhA is calculated from raw data on R and Xc (PA [degrees] = arctangent (Xc/R) x (180°/π)), and it is considered one of the indicators of cellular health and cell membrane structure18, as well as an indicator of the distribution of ICW and ECW spaces, i.e., altered redistributions of the compartments (specifically, changes from intracellular to extracellular water, which low phase angles can show)19. Thus, a low PhA value may be due to overhydration and/or malnutrition, and the Z-score could be used to differentiate if this low PhA is due to the loss of soft tissue mass, an increase in tissue hydration, or both. In addition, the transformation of the Z-score could help researchers compare changes in different study populations3,14.
In addition, PhA is considered a predictor of health status, an indicator of survival, and a prognostic marker for different clinical outcomes3,20, even under other clinical conditions20,21,22,23, where high PhA values indicate greater cell membrane integrity and vitality10,13and therefore greater functionality. This is in contrast to low PhA values, which reflect membrane integrity and permeability loss, leading to impaired cell function or even cell death14,22,24. In patients with chronic heart failure (CHF), smaller PhA values were associated with a worse functional class classification25. In addition, one of the advantages of PhA measurement is that it does not require recalled parameters, body weight, or biomarkers.
Several studies have recommended the use of raw BIA measurements in patients who had alterations in fluid shifts and fluid redistributions or non-constant hydration status, such as those in AHF26. This was because BIA is based on regression equations that estimate total body water (TBW), extracellular body water (ECW), and intracellular body water (ICW). Therefore, the lean and fat mass estimations in such patients are biased because of the physiological relationship with soft tissue hydration27.
The bioelectrical impedance vectorial analysis (BIVA) method overcomes some limitations of the conventional BIA method28. It provides additional information through a semiquantitative evaluation of body composition in terms of body-cell mass (BCM), cell mass integrity, and hydration status29. Thus, it allows an estimation of the body fluid volume through vector distribution and distance patterns on an R-Xc graph28,30. BIVA is used to create a vector plot of impedance (Z) using the whole-body R and Xc values derived from BIA at a frequency of 50 kHz.
To adjust the raw values of R and Xc, the parameters R and Xc are standardized by height (H), expressed as R/H and Xc/H in Ohm/m, and plotted as a vector; this vector has a length (proportional to the TBW) and a direction on the R-Xc graph16,28.
A sex-specific R-Xc graph contains three ellipses, which correspond to the 50%, 75%, and 95% tolerance ellipses of a healthy reference population28,31,32; the ellipsoidal form of the ellipses is determined by the relationship between R/H and Xc/H. However, to evaluate the impedance parameters in a gender-specific reference health population, the original raw BIA parameters were transformed into bivariate Z-scores (in a BIVA Z-score analysis) and plotted on an R-Xc Z-score graph33,34. This graph, compared with an R-Xc graph, represented the standardized R/H and Xc/H as a bivariate Z-score, i.e., Z(R) and Z(Xc) showed the number of standard deviations away from the mean value of the reference group33. The tolerance ellipses of the Z-score conserved the same pattern as that of the ellipses for the percentiles on the original R-Xc graph31,33. The Z-score graphs for R-Xc and R-Xc showed changes in soft tissue mass and tissue hydration independent of regression equations or body weight.
Vector displacements along the major axis of the ellipses indicated changes in hydration status; a shortened vector that fell below the 75% pole of an ellipse indicated pitting edema (sensibility = 75% and specificity = 86%); however, the optimal threshold for the detection of pitting edema was different in AHF and CHF patients, where the lower pole of 75% corresponded to AHF patients, and 50% corresponded to CHF patients edema (sensibility = 85% and specificity = 87%)35. On the other hand, vector displacements along the minor axis corresponded to cell mass. The left side of the ellipses indicated a high cell mass (i.e., more soft tissue), where shorter vectors corresponded to obese individuals and were characterized by phases similar to those of athletic ones, who had longer vectors. On the contrary, the right side indicated less body cell mass21,34; according to Picolli et al.31,33, the scores of the vectors of the anorexia, HIV, and cancer groups were located on the right side of the minor axis, which corresponds to the category of cachexia.
This study aimed to explain how to obtain and interpret PhA values by using BIA in patients with AHF who were admitted to an emergency department and to show their clinical applicability/usefulness as a predictive marker for the prognosis of 90-day events.
The protocol was approved by the Research Ethics Committee of the National Institute of Medical Sciences and Nutrition Salvador Zubirán (REF. 3057). To conduct BIA measurements, tetrapolar multiple-frequency equipment was used (see Table of Materials). This equipment provided accurate raw values for the resistance (R), reactance (Xc), and phase angle (PhA) at a frequency of 50 kHz, which allowed the impedance to be measured with the best signal-to-noise ratio. The adhesive electrodes used needed to correspond to the manufacturer's recommendations. Informed written consent was obtained from the patients involved in the study.
1. Experimental and patient preparation
NOTE: These steps were performed before performing a BIA measurement.
2. BIA measurement
3. Analysis of BIA raw parameters on the R-Xc Z-score graph
4. Interpretation and analysis of the BIVA Z-score
NOTE: Identify the four patterns in the R-Xc Z-score graph. In the extremes along the major axis, the lower pattern is associated with congestion, whereas the upper pattern is associated with dehydration status. In the extremes along the minor axis, the left pattern is associated with more cell mass in soft tissues, whereas the right pattern is associated with less cell mass in soft tissues. To calculate the bivariate Z-score from the mean age of the group, the following formula is used: Z(R) = (R/H mean age group – R/H mean value in the reference population) / standard deviation of the reference population and Z(Xc) = (Xc/H mean age of the group – Xc/H mean value in the reference population) / standard deviation of the reference population.
5. Directly calculating and interpreting PhA
NOTE: Raw R50 and Xc50 values are needed to calculate PhA.
According to the protocol described above, we present data from four AHF patients (two females and two males) who were admitted to an emergency department as an example of the clinical applicability of phase angle values and BIVA Z-score analysis. BIA measurements were performed using phase-sensitive multiple-frequency equipment within 24 h of admission.
To calculate the bivariate Z-score from the mean of the age group, the following formula was used: Z(R) = (R/H mean value of the age group – R/H mean value of the reference population) / standard deviation of the reference population, and Z(Xc) = (Xc/H mean value of the age group – Xc/H mean value of the reference population) / standard deviation of the reference population.
After the BIA measurements, the patients were classified according to their PhA values at admission into two categories: (1) PhA < 4.8° and (2) PhA ≥ 4.8°. An event was defined if the patient presented in-hospital mortality, out-of-hospital mortality, or re-hospitalization for any cause within 90 days after discharge. The clinical characteristics of the patients are presented in Table 1, and Table 2 shows the laboratory and echocardiographic characteristics of the two men and two women-divided according to PhA- upon admission.
Case 1 corresponded to a 75-year-old woman without a previous diagnosis of HF who was admitted due to edema and dyspnea with one month of evolution after a hip surgery that took place two months earlier. On arriving, she had Godet edema (+++), rales, and S3 sound, which were reported. The imaging findings were vascular congestion (predominantly right bilateral pleural effusion); she also presented hypoalbuminemia, hyperphosphatemia, type I respiratory failure, and a wet-warm hemodynamic profile of acute heart failure according to the European Society of Cardiology (ESC) Guidelines40. Based on the PhA and BIVA Z-score analysis (Figure 1; Group 1), the patient had tissue congestion with a loss of BCM related to malnutrition, which was consistent with the systemic inflammatory episode because the increased hydrostatic and oncotic pressures that were involved caused the leakage of fluid into the interstitial space. The patient presented with an event (re-hospitalization) 11 days after hospital discharge.
Case 2 referred to an 83-year-old woman with CHF and reduced left ventricular fraction ejection (LVEF) who was admitted due to dyspnea within 7 days of evolution and did not develop edema or rales. According to the BIVA Z-score analysis (Figure 1; Group 2), the patient was within the limits of the 75% tolerance ellipse in a non-congestion area, which reflected a dry profile that indicated no tissue or intravascular congestion. In addition, despite the patient's advanced age, the BCM was preserved, in addition to a PhA of 5.4°, which demonstrated good cellular vitality. These characteristics were concordant with the patient's evolution, as no events were presented.
Case 3 corresponded to a 78-year-old man who was admitted due to progressive edema associated with a decreased functional class and dyspnea. On admission, he had Godet edema (+++), and a chest X-ray revealed fluid overload, cardiomegaly, and predominantly left bilateral pleural effusion without any infectious processes, which reflected a wet-warm clinical profile. The BIVA Z-score (Figure 2, Group 3) and the PhA of 2.5° showed that the patient had tissue congestion, as in Case 1; there was a redistribution of fluids due to the increased hydrostatic and oncotic pressures. He died three days after hospital admission.
Case 4 corresponded to an 80-year-old man with chronic heart failure and reduced LVEF who was admitted due to dyspnea within 6 days of evolution; he did not develop edema or rales. An X-ray showed interstitial thickening and a prominent aortic arch. According to the BIVA Z-score (Figure 2; Group 4), the patient had no congestion, and the albumin levels were normal; thus, an imbalance between the hydrostatic and oncotic pressures was avoided. However, the displacement vector to the right reflected the loss of soft tissue. As in Case 2, the patient did not present an event.
The results show that patients who were classified with congestion, PhA < 4.8°, and less BCM according to the BIVA Z-score analysis had poor prognoses that were related to other predictors, such as the length of stay, serum albumin, and brain natriuretic peptides.
Figure 1: R-Xc z-score graph with data of AHF female patients admitted to the emergency department. The figure reflects the two female patients, and both vectors fell below 75% tolerance ellipses in the water increase quadrant (congestion status). Group 1 corresponds to the vector of Case 1, and Group 2 corresponds to the vector of Case 2. Please click here to view a larger version of this figure.
Figure 2: R-Xc z-score graph with data of AHF male patients admitted to the emergency department. The figure reflects the two male patients, the vector fell below 75% tolerance ellipses (congestion status) and corresponds to Case 3 (Group 3), and the vector classified in the non-congestion area corresponds to Case 4 (Group 4). Please click here to view a larger version of this figure.
Table 1: Characteristics of the patients at emergency department admission according to phase angle upon admission. BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure; LOS: length of stay. Please click here to download this Table.
Table 2: Laboratory results at emergency department admission and echocardiographic characteristics according to phase angle upon admission. SaO2: Oxygen saturation; PaO2: Partial pressure of oxygen; PaCO2: Partial pressure of carbon dioxide; HCO3: Bicarbonate; FS: fractional shortening; LVEF: left ventricular ejection fraction. Please click here to download this Table.
This protocol describes the utility of using R-Xc Z-score analysis in clinical practice for patients admitted to an emergency department with AHF. Considering that in patients with AHF, the main reason for hospital admission is congestion, its quick and accurate detection, and evaluation are crucial for patients' outcomes6.
This article illustrates the variety of clinical manifestations of AHF and how BIVA Z-score analysis (congestion status and BCM) can be used to accurately and reliably evaluate and classify patients; in addition, the characteristics of the patients with PhA <4.8° were consistent with other predictors that have been associated with poor prognoses, such as low serum albumin levels, greater lengths of stay, and higher brain natriuretic levels35.
An R-Xc Z-score graph can be used to evaluate congestion status and BCM. Hence, the implementation of the PhA, in addition to the R-Xc Z-score graph, provides useful and accurate information during the evaluation of congestion; it is also a diagnostic tool for assessing the presence of subclinical congestion and clinical congestion and peripheral edema41. In addition, it can serve as a monitoring tool since minimal changes in hydration and nutritional status are detectable in patients with acute and chronic HF during hospitalization5,21; finally, it can serve as a predictor of poor outcomes. Furthermore, variations in the values are due to alterations in fluid and nutritional status39. Additionally, when combined with biomarkers and clinical judgment, it can help drive physicians' decisions on effective diuretic therapeutic strategies and the management of AHF patients10.
Several studies have shown that the PhA is an independent prognostic marker of poor prognosis in AHF42 and CHF, regardless of whether patients have right or left HF21,43. In the literature, it has been reported that PhA decreases in patients with edema and fluid retention5, as well as in patients with functional classes III-IV from the New York Heart Association (NYHA)25, which was consistent with the present results. Nevertheless, PhA increases after clinical stabilization of a patient21,22. The results that we observed were similar to those found by Alves et al.20, who showed that a PhA of <4.8° was a predictor of mortality over an average follow-up period of 24 months (sensibility = 85% and specificity = 45%; AUC: 0.726); additionally, this cut-off point was found to be a predictor of in-hospital mortality and re-hospitalization within 90 days after discharge39. It is important to recognize that multiple studies have reported different cut-off points for PhA with different outcomes in HF patients. Scicchitano et al.44 demonstrated that a PhA of ≤4.9° independently predicted all-cause death (sensibility = 75%, and specificity = 44%); Massari et al.35 found that even in AHF and CHF, peripheral fluid accumulation significantly decreased PhA (4.2° vs. 4.5°, respectively); Colín et al.22 found that in outpatients with CHF, a PhA of <4.2° was a predictor of mortality at 3 years for deaths from all causes (HR: 3.08, 95%IC: 1.06-8.99).
To the best of our knowledge, only one previous study by Piccoli41 evaluated BIVA Z-scores to determine patients with acute dyspnea of cardiac or noncardiac origin; however, the strength of this work is its evaluation of AHF patients with the BIVA Z-score in conjunction with the PhA in relation to patients' prognoses.
The advantages of PhA are that it does not require measurements of body weight and/or height, and it cannot be influenced by the presence and activity of a pacemaker (PM) or implanted cardioverter defibrillator (ICD)44,45,46.
Technical concerns: accuracy of the device, agreement, and types of electrodes
A crucial requirement is to use a phase-sensitive device to ensure the reliable and accurate evaluation of PhA values and hydration. The device accuracy is evaluated using a high-precision (<1%) circuit consisting of a resistor and a capacitor connected in parallel16. Also, excellent intra-observer repeatability for R, Xc, and PhA has been determined47.
The PhA can be obtained from single-frequency (SF) or multi-frequency (MF) devices. The intra-observer repeatability in R50, Xc50, and PhA50 is high; however, the agreement of the PhA values between these devices is questionable47,48. The poor correlation between SF-frequency and MF-frequency devices does not affect the classification of the hydration status or BCM (quadrants or categories); it is necessary to take caution in interpretation because minimal differences (<0.5°) can be used to discriminate between healthy and critical patients13 due to the underestimation PhA and Xc in CHF patients with MF-BIA47.
Due to the absence of international manufacturing standards, the cross-calibration of different instruments' electrical accuracy is essential for impedance companies14; in addition, the electrodes to be used are from a manufacturer's equipment. Nevertheless, even ideally, every Ag/AgCl electrode should have the same intrinsic impedance, and there should be differences between the electrodes. Nescolarde et al.49 observed a large variability of the intrinsic R (11-665 Ω) and Xc (0.25-2.5 Ω) values among nine types of electrodes that were composed of silver-silver chloride (Ag/AgCl). This systematically and significantly affected the vector length and position on R-Xc graph and consequently affected the PhA values.
The perspectives of PhA include the evaluation of the percentage of change or its absolute delta (Δ) in order to determine the optimal changes or even the velocity of this change after clinical stabilization as a biomarker for verifying the response to treatment or therapy.
The authors have nothing to disclose.
The authors would like to thank Prof(s). Piccoli and Pastori of the Department of Medical and Surgical Sciences, University of Padova, Italy, for providing the BIVA software. This research did not receive any specific grant from funding, agencies in the public, commercial, or not-for-profit sectors. This protocol/research is part of the Ph.D. dissertation of María Fernanda Bernal-Ceballos supported by the National Council of Science and Technology (CONACYT) scholarship (CVU 856465).
Alcohol 70% swabs | NA | NA | Any brand can be used |
BIVA software 2002 | NA | NA | Is a sofware created for academic use, can be download in http:// www.renalgate.it/formule_calcolatori/ bioimpedenza.htm in "LE FORMULE DEL Prof. Piccoli" section |
Chlorhexidine Wipes | NA | NA | Any brand can be used |
Examination table | NA | NA | Any brand can be used |
Leadwires square socket | BodyStat | SQ-WIRES | |
Long Bodystat 0525 electrodes | BodyStat | BS-EL4000 | |
Quadscan 4000 equipment | BodyStat | BS-4000 | Impedance measuring range: 20 – 1300 Ω ohms Test Current: 620 μA Frequency: 5, 50, 100, 200 kHz Accuracy: Impedance 5 kHz: +/- 2 Ω Impedance 50 kHz: +/- 2 Ω Impedance 100 kHz: +/- 3 Ω Impedance 200 kHz: +/- 3 Ω Resistance 50 kHz: +/- 2 Ω Reactance 50 kHz: +/- 1 Ω Phase Angle 50 kHz: +/- 0.2° Calibration: A resistor is supplied for independent verification from time to time. The impedance value should read between 496 and 503 Ω. |