Application of multi-frequency cardiometabolic risk bioimpedance - - PDF document

▶
application of multi frequency
SMART_READER_LITE
LIVE PREVIEW

Application of multi-frequency cardiometabolic risk bioimpedance - - PDF document

Dec 31, 2022 518 likes •614 views

Beyond BMI and Waist Circumference in obesity management and Application of multi-frequency cardiometabolic risk bioimpedance analysis to the Reduction in fat mass management of patients with Maintenance of lean body mass (fat-free

slide-1
SLIDE 1

Application of multi-frequency bioimpedance analysis to the management of patients with

  • besity and metabolic disorders

Lindsay Plank

Department of Surgery University of Auckland Auckland, New Zealand

  • Reduction in fat mass
  • Maintenance of lean body mass (fat-free mass)
  • Reduction of central fat deposition, esp. visceral fat
  • BMI is uninformative for these aspects of body

composition

  • BMI cut-offs for overweight/obesity are problematic in

non-European populations

  • WC does not identify visceral fat
  • Bioimpedance analysis: a tool for evaluating fat

mass, skeletal muscle mass, visceral fat, their changes with weight loss, and their relationships to cardiometabolic disorders?

Beyond BMI and Waist Circumference in

  • besity management and

cardiometabolic risk

Percent fat vs BMI

Rush, Freitas, Plank Br J Nutr 2009;102: 632

Women

BMI, kg/m2 15 20 25 30 35 40 45 50 Body fat, % 10 20 30 40 50 60

AI E PI M

Percent fat vs BMI

Obese

Impedance (derived from measured voltage)

~

Constant alternating current

Bioelectrical impedance devices measure impedance at one

  • r more frequencies

They don’t ‘measure’ water volumes or fat mass. These must be estimated in some way from impedance values

Impedance Constant current

Hand-to-foot Standing

  • r

supine Leg-to-leg Standing (or arm-to-arm)

Basic principle

For a cylindrical conductor full of electrolyte: Impedance Z = ρL/A (ρ = electrolyte resistivity) i.e, Z = ρL2/(AL) and volume of electrolyte = AL

Length x Length Impedance Electrolyte volume = ρ Height x Height Impedance, Z Total fluid volume ∝ (impedance index) Human body is modelled as a single cylinder Conductor length ~ 95% Height

Stahn et al Handbook of Anthropometry 2012

slide-2
SLIDE 2

To increase the strength of the relationship between the impedance index and fluid volume: weight, age, sex added as predictors Typically, a gold-standard determination of total body water allows an equation to be developed for predicting this volume from H2/Z, weight, age, sex

Total body fat Fat-free mass Water Mineral

Carbohydrate

Protein

73% of FFM H2/Z, weight, age, sex will predict FFM as measured by UWW, DXA or 4-C model, allowing TBF estimation A large number of equations have been published - they tend to be specific for the populations on which they were developed

ECW ICW Water

RECW CM RICW

Impedance (Z) consists of both resistive (R) and reactive (X, cellular tissue) components Cell and tissue interfaces act like capacitors and their impedance to current flow (termed reactance) is frequency dependent

Stahn et al Handbook of Anthropometry 2012

HF current LF current

Cells ECW ICW

Capacitors present very high impedance to low frequency current – at zero frequency no current through cells and Z0 = RECW (determination of ECW volume) At high frequencies, current penetrates cells – at infinite frequency, Z∞ = parallel combination of RECW and RICW (determination of TBW volume) At intermediate frequencies, measured impedance is a combination of resistance (R) and reactance (X) Reactance causes the current to lag behind the voltage creating a phase shift:

Magnitude (mV, mA) Time Current Voltage

phase shift

  • can be calculated as an angle

φ = arctan (X/R) Phase angle

X R Z

Bioelectrical Impedance Analysis

Single frequency (SFBIA) Multi-frequency (MFBIA) Bioimpedance spectroscopy (BIS)

Typically 50kHz Empirical eqns Typically 2-7 selected freqs Empirical eqns

Increasing complexity, device size, cost

Model-fitting Less population specific

Single Frequency BIA

Most widely used approach since development of the technology in 1979 by RJL Systems. In addition to TBW and FFM estimation, the close relationship between ECW and TBW in health allows ECW estimation. Originally developed for supine measurement, standing leg-to-leg devices have proliferated. Contribution to Z ~50% ~45% ~5%

H2/Z highly correlated for whole body, legs and arms. Leg-to-leg and arm-to-arm measurements have similar predictive ability as whole body. (~50% total mass)

slide-3
SLIDE 3

2

Nunez C et al Med Sci Sports Exer 1997;29:524

Comparative evaluation of leg-to-leg and traditional hand-to-foot devices H2/Z

DXA FFM (kg)

H2/Z leg-to-leg traditional

r = 0.89 SEE = 6.1kg r = 0.95 SEE = 4.3kg

(Addition of age and sex to regression models improved R2 to similar values)

Recommendations for reliable supine whole-body measurements

  • Conditions should be same as those under which the equations
  • r model used were developed
  • Check device calibration regularly
  • Fast (except water) and avoid alcohol, caffeine and exercise for

at least 8 h

  • Empty bladder before measurement
  • Swab skin with alcohol
  • Proximal electrode positioning is critical
  • Avoid excessively warm or cold ambient temperature
  • Arms ~30 deg from trunk; legs ~45 deg separation (limbs

straight); towels/blankets for obese

  • Consistency vital for serial measurements (time of day)
  • Supine for at least 10 min before measurement

(follow manufacturer’s instructions – Quadscan manual suggests 3-4 min)

Shirreffs & Maughan Eur J Appl Physiol 1994;69:461

Standardization crucial 13 Ω increase over 1 h – translates to 500mL reduction in TBW (Z measured supine after standing for 1 h)

Time (min) Change in Z (Ω) Time (min)

Raw impedance data (resistance, reactance) at 50 kHz provide useful information: Phase angle (PA) Bivariate vector analysis (BIVA)

X R

φ

Z Phase angle

(independent of regression equations)

Phase angle

Depends on X which is a measure of cellular mass and cellular integrity Depends on R which is a measure of fluid status Low PA occurs in malnutrition and poor cellular function Low PA (at 50 kHz) predicts poor survival in liver disease, HIV/AIDS, lung, breast, colorectal, pancreatic cancer

Liver cirrhosis

Selberg & Selberg Eur J Appl Physiol 2002;86:509

Multi-frequency BIA

Bodystat Quadscan 4000 Impedance measured at: 5, 50, 100, 200 kHz Prediction Marker, PM = Z200 Z5 ECW TBW ∝ FFM estimated from 50 kHz impedance data Increasing PM generally indicates poorer health (fluid retention)

slide-4
SLIDE 4

Type 2 diabetics undergoing bariatric surgery

Body composition

GE-Lunar iDXA

Male : Female 32 : 37 Age (y) 47.0 ± 6.9 (20 − 56) Weight (kg) 116.2 ± 19.4 (82.0 − 165.7) BMI (kg/m2) 40.0 ± 6.0 (28.5 − 60.4)

mean ± SD (range)

Bodystat Quadscan 4000

Pre-surgery characteristics

Baseline 12 months 10 20 30 40 50 60 70

Weight (kg) BMI Fat (kg) %fat

Baseline 12 months 20 40 60 80 100 120 140 160

Wgt

Baseline 12 months 10 20 30 40 50 Baseline 12 months 10 20 30 40 50 60 Baseline 12 months 20 40 60 80 100

FFM (kg)

DXA BIA DXA BIA DXA BIA

Baseline 12 months 2 4 6 8

Women Men

PA (deg)

Changes: W: P<0.0001 M: P=0.002

Baseline 12 months 10 20 30 40

Women Men

ICW (L)

Changes: W: P<0.0001 M: P=0.0001

PM

Baseline 12 months 0.0 0.2 0.4 0.6 0.8 1.0

Women Men

Changes: W: P<0.0001 M: P=0.005

Baseline 12 months 5 10 15 20 25 30

ECW (L)

Changes: W: P<0.0001 M: P<0.0001

Baseline 12 months 10 20 30 40 50 60 70

TBW (L)

Changes: W: P<0.0001 M: P<0.0001

30 40 50 60 70 80 90 100

  • 15
  • 10
  • 5

5 10 15 20 20 30 40 50 60 70 80 90 100 20 40 60 80 100

Baseline Body Fat Women Baseline Body Fat Men

20 30 40 50 60 70 80 90 100 20 40 60 80 100

BIA BIA DXA DXA Average DXA and BIA BIA minus DXA

mean 2SD

  • 2SD

BIA minus DXA

mean 2SD

  • 2SD

Average DXA and BIA

20 40 60 80 100

  • 14
  • 12
  • 10
  • 8
  • 6
  • 4
  • 2

2

r=0.958 r=0.975

  • 6.4
  • 13.5%
  • 0.8
  • 1.6%
  • 12.0
  • 25%

0.2 0.4%

  • 7.6
  • 13.9%

8.0 14.0%

12 month Body Fat Women 12 month Body Fat Men

10 20 30 40 50 10 20 30 40 50 5 10 15 20 25 30 35 40 45

  • 12
  • 10
  • 8
  • 6
  • 4
  • 2

2 4 10 20 30 40 50 60 10 20 30 40 50 60 10 20 30 40 50 60

  • 8
  • 6
  • 4
  • 2

2 4 6 8 10 12

BIA BIA DXA DXA BIA minus DXA

mean 2SD

  • 2SD

BIA minus DXA

mean 2SD

  • 2SD

Average DXA and BIA Average DXA and BIA

r=0.966 r=0.931

0.2 0.7%

  • 4.9
  • 16.2%

5.3 17.5%

  • 4.1
  • 16.8%

1.8 7.2%

  • 10.0
  • 41%
slide-5
SLIDE 5

Body Fat Changes over 12 months - Women Body Fat Changes over 12 months - Men

10 20 30 40 50 60

  • 15
  • 10
  • 5

5 10 15 20 10 20 30 40 50

  • 8
  • 6
  • 4
  • 2

2 4 10 20 30 40 50 60 10 20 30 40 50 60 10 20 30 40 50 60 10 20 30 40 50 60

ΔBIA ΔBIA Δ DXA ΔDXA ΔBIA minus ΔDXA

mean 2SD

  • 2SD

ΔBIA minus ΔDXA

mean 2SD

  • 2SD

Average ΔDXA and ΔBIA Average ΔDXA and ΔBIA

r=0.936 r=0.977

0.0 0%

  • 8.3
  • 35%

8.3 35%

  • 2.2
  • 9.5%

2.1 9.3%

  • 6.5
  • 28%

FM (kg) BIA minus ADP

Body composition

Bodystat Quadscan 4000

Clin Nutr 2008; 27: 350

Average ADP and BIA Average ADP and BIA Air Displacement Plethysmography (ADP) Mean BMI 45.2 (n=83) Mean BMI 49.9 (n=36)

%BF: BIA vs DXA or 3C model in women undergoing obesity surgery

Pre-surgery Post-surgery Bias Limits of agreeement Auckland Auckland Das et al* Das et al 0.3

  • 6.0, 6.6

0.1

  • 7.0, 7.2

1.1 * RJL 50 kHz, Lukaski equations (J Appl Physiol 1986)

Das SK et al Am J Physiol 2003; 284: E1080

  • 4.7, 6.9
  • 10.6, 11.4

0.4 Change

  • 10.0, 9.3
  • 0.4
  • 0.3
  • 8.8, 8.2

WC>88cm Obesity 2006;14:1731–1738 %BF BC418 minus DXA

Savastano S et al Obes Surg 2010; 20: 332

Morbidly obese women undergoing bariatric surgery

Pre-surgery 12 months post-surgery Fat Mass (kg) BIA minus DXA Average DXA and BIA (kg) Mean BMI 42.1 Average DXA and BIA (kg) Mean BMI 33.4 BIA device: RJL BIA101 with manufacturer’s equations

17%

  • 13%

17%

  • 8%

Am J Clin Nutr 2005; 81: 74

491 women, 100 men (BMI: 17 – 55) DXA vs Quadscan 4000 for %BF (so single-frequency analysis)

women men

%BF by DXA %BF by DXA %BF DXA minus BIA

BIA underestimates %BF at higher adiposity

slide-6
SLIDE 6

Obesity and overhydration

Shen et al Obes Res 2003;11:5

~80% fat, 14% water ECW/ICW ~3.7 Adipose Tissue ECW/ICW Non-obese 0.63 Obese 0.81

Waki et al AmJ Physiol 1991;261:E199

TBW/FFM Non-obese 0.73 Obese 0.76 Post GBP 0.75

Das et al Am J Physiol 2003;284:E1080

Body fat estimation based on TBW

Quadscan determination of TBW (200 kHz) Assume FFM = TBW/0.76 Pre-bariatric surgery: Assume FFM = TBW/0.75 Post-bariatric surgery:

Baseline Body Fat Women Baseline Body Fat Men

Average DXA and BIA

mean 2SD

  • 2SD

mean 2SD

  • 2SD

Average DXA and BIA Average DXA and BIA BIA minus DXA BIA minus DXA Average DXA and BIA

30 40 50 60 70 80 90 100

  • 15
  • 10
  • 5

5 10 15 20 20 40 60 80 100

  • 14
  • 12
  • 10
  • 8
  • 6
  • 4
  • 2

2 30 40 50 60 70 80 90 100

  • 10
  • 5

5 10 15 20

  • 13.5%
  • 1.6%
  • 25%

0.4%

  • 13.9%

8.0 0.2

  • 7.6

14.0% 7.5 0.0

  • 7.5

0.0%

  • 13.9%

13.9%

20 40 60 80 100

  • 12
  • 10
  • 8
  • 6
  • 4
  • 2

2 4 6

  • 6.4
  • 0.8
  • 12.0
  • 3.1

4.0

  • 10.2
  • 6.3%

8.1%

  • 21%

From FFM (50kHz) From TBW (200kHz)

12 mo Body Fat Women 12 mo Body Fat Men

Average DXA and BIA

mean 2SD

  • 2SD

mean 2SD

  • 2SD

Average DXA and BIA Average DXA and BIA BIA minus DXA BIA minus DXA Average DXA and BIA From FFM (50kHz) From TBW (200kHz)

10 20 30 40 50 60

  • 8
  • 6
  • 4
  • 2

2 4 6 8 10 12

0.2 0.7%

  • 4.9
  • 16.2%

5.3 17.5%

5 10 15 20 25 30 35 40 45

  • 12
  • 10
  • 8
  • 6
  • 4
  • 2

2 4

  • 4.1
  • 16.8%

1.8 7.2%

  • 10.0
  • 41%

0.2

  • 4.9

5.3

10 20 30 40 50 60

  • 8
  • 6
  • 4
  • 2

2 4 6 8 10 12

  • 0.4
  • 4.4

3.6

  • 1.3%
  • 14.7%

12.0%

10 15 20 25 30 35 40 45 50

  • 10
  • 8
  • 6
  • 4
  • 2

2 4 6

  • 1.9

3.6

  • 7.4
  • 7.3%

13.9%

  • 29%

Summary points

  • Limited data in obese comparing single-frequency

BIA fat mass estimation with reference data

  • As a consequence, almost no development of BIA

equations suitable for obese populations

  • Obesity is a challenge for BIA given the prevalence of

truncal adiposity and variability in hydration

  • Acceptable accuracy achievable in women; limits of

agreement wide

  • DXA as a reference technique also has limitations

Bioimpedance spectroscopy

RECW CM RICW

Cole equivalent circuit model

Stahn et al Handbook of Anthropometry 2012

HF current LF current

Cells ECW ICW R0=RECW R∞=RECW//RICW

R50 X R

Z

Increasing frequency

R0 R∞ φ

Cole plot

50 kHz

An extension of the multi-frequency approach - fit measured data to the Cole plot,

  • btain R0 and R∞ by

extrapolation.

slide-7
SLIDE 7

Hanai Mixture Theory for Fluid Volumes

  • Describes effect on conductivity of conductive

medium with non-conductive material in suspension

  • Volume (ECW, TBW) estimation requires:

resistivities of ECW and ICW body density form factor (accounts for geometrical properties

  • f arms, leg, trunk)
  • Accuracy of estimation dependent on choosing

correct parameters – may differ for different populations (estimated using regression techniques)

  • Same drawbacks as regression-based approaches?

Impedimed SFB7 256 freqs 4 kHz – 1MHz Xitron 4000B 50 freqs 5 kHz – 1MHz Bodystat Multiscan 5000 50 freqs 5 kHz – 1MHz

Bioimpedance spectroscopy

  • FFM estimated from TBW using appropriate

hydration factor (eg 0.732)

  • Emphasis on fluid volume estimation in published

literature

  • ICW from TBW and ECW – a measure of body cell

mass (and skeletal muscle)

  • BCM contains all the metabolically active tissues of

the body (including muscle cells, organ cells, blood cells, and immune cells) and its fluid content is ICW

Clin Nutr 2008;27:832

Quadscan 4000 and Xitron Hydra 4200 vs isotope dilution (deuterium, bromide) for ICW as measure of BCM.

Baseline 12 months 10 20 30 40

Women Men

ICW (L)

Changes: W: P<0.0001 M: P=0.0001

Baseline 12 months 20 40 60 80 100

FFM (kg)

DXA BIA

Changes: DXA: P<0.0001 BIA: P<0.0001

Pre- and post-surgery – Quadscan estimation of FFM and ICW Reduction in ICW consistent with a reduction in skeletal muscle leading to a reduction in FFM. Quadscan estimates FFM and ICW by independent approaches.

Damms-Machado et al Obes Surg 2012;22:881

Nutriguard-M MFBIA (5, 50, 100kHz) Body composition before and after sleeve gastrectomy

slide-8
SLIDE 8

Visceral fat estimation

J Hum Nutr Diet 2013;26 (Suppl. 1):154

Omron BF-500 body composition monitor Standing 8-electrode device Provides VFAT in 30 levels BF-500 vs Bodystat 1500 for fat mass WC vs VFAT level VFAT > 10 pts yielded sensitivity of 100% and specificity of 82% for predicting MS, correctly classifying 94% of MS cases.

Tanita AB-140 ViScan compared to MRI for abdominal fat compartments

Thomas E L et al Eur J Clin Nutr 2010;64:525

ViScan visceral fat (arbitrary units) ViScan trunk fat (%) MRI total abdominal adipose tissue (% of body weight) MRI Intra-abdominal adipose tissue volume (L)

r2 = 0.88 r2 = 0.53 Thomas E L et al Eur J Clin Nutr 2010;64:525

ViScan visceral fat rating ViScan trunk fat rating

MRI total abdominal adipose tissue (% of body weight) MRI intra-abdominal adipose tissue volume (L)

Same ViScan visceral fat scores. Similar total abdominal adiposity by MRI

MRI intra-abdominal adipose tissue ~2x higher (and less subcutaneous adipose tissue)

ViScan predicts total abdominal adiposity but visceral fat prediction limited. Can it predict changes in abdominal fat compartments?

Tanita AB-140 ViScan (n=285 volunteers) Visceral fat measurement was similar to WC for predicting metabolic syndrome – may offer objectivity for repeat measurements and reduced inter-observer variability compared to WC

OMRON HDS-2000 DUALSCAN system

Shiga T, et al. IFMBE Proc 2007;17:687

67 patients with obesity and obesity- related co-morbidities placed on calorie restriction

CT intra-abdominal AT area (cm2) Dual-BIA intra-abdominal fat area (cm2)

Ida M et al. Obesity 2013;21:E350

Change from baseline (%))

(no VFAT verification by repeated CT)

Segmental BIA

Stahn et al Handbook of Anthropometry 2012

slide-9
SLIDE 9

X/H, Ohm/m

10 20 30 40 50 60 70 100 200 300 400 500 600 700

BIVA: Resistance-Reactance graph

Piccoli A Contrib Nephrol 2005;149:150

95% tolerance 75% tolerance 50% tolerance 10 20 30 40 50 60 70 100 200 300 400 500 600 700

R/H, Ohm/m R/H, Ohm/m

Females

10 20 30 40 50 60 70 100 200 300 400 500 600 700

R/H, Ohm/m Xc/H, Ohm/m Pre-surgery Post-surgery

Males

10 20 30 40 50 60 70 100 200 300 400 500 600 700

R/H, Ohm/m Xc/H, Ohm/m Pre-surgery Post-surgery

Females

10 20 30 40 50 60 70 100 200 300 400 500 600 700

R/H, Ohm/m Xc/H, Ohm/m

Pre- to 12 months post-surgery trajectories for 9 women Pre- to 12 months post-surgery trajectories for 8 women Females

10 20 30 40 50 60 70 100 200 300 400 500 600 700

R/H, Ohm/m Xc/H, Ohm/m

Summary

BIA equations developed for use in the obese are lacking This likely includes manufacturers’ proprietary equations Future work on segmental BIA approaches may improve agreement with DXA (or multi-compartment models) for fat mass Usefulness of visceral fat by BIA is currently questionable Bivariate Vector Analysis may offer useful qualitative information for tracking individuals