P.72 UK
Indications for home enteral tube feeding in the
AI in mixtures (#mol I-1)
AI PN intakes* (#mol kg-1 d-l)
Plasma AI** (#molI-1 )
3.4_+0.6 1.4-+0.9
0.3+0.1 0.08_+0.03
2.0_+0.4 0.9_+0.5
British Artificial Nutrition Survey (BANS) (BAPEN), M. Ella (chairman for correspondence) Dunn Clinical Nutrition Centre, Cambridge CB2 2DH, UK.
Study (a) (1990) n = 8 Study (b) (1998) n= 10
Home enteral tube feeding (HETF) in the UK has grown by over 20% in one year, but there is little information about the indications and outcome. The aim of this study was to undertake a prospective national survey to provide baseline data on the major indications and age distribution of adult patients requiring HETE Information was obtained using 5779 individual patient questionnaire forms from 219 centres between March 1996 and December 1997. They were completed by health professionals, mainly dietitians, when patients were discharged into the community. Sixty per cent were aged over 65 years (35% over 75y), 15% were aged 55-65 y and 24% 16-54 y.
The commonest overall indication (all ages) was cerebrovascular accident, which accounted for 37% of all indications and 62% in those over 75y. The commonest indication (33%) in those aged 55-64y was carcinoma of the oropharynx plus oesophagus; (23% oropharynx; 10% oesophagus), and in those aged 16-54 y it was multiple sclerosis (21%). This study shows that more than half of the HETF provided to adults in the UK between 1996 and 1997 involved elderly subjects over 65y, and that more than half of these suffered a cerebrovascular accident. Outcome data in the various groups of patients indicated in the table are urgently required.
Session 1 - TOPIC 9: N U T R I T I O N A L A S S E S S M E N T II The aim of the present study was to determine in both sexes the relationships between BIA and some common individual characteristics. Single-frequency BIA was determined according to standardized procedures (after voiding, early in the morning, after lying in the supine position for 20 min, at ambient temperature of 22-24°C) in a cohort of lean and overweight subjects belonging to a healthy population (18-80y): 1047 men [mean: age 49.0y, Ht 171cm, Wt 70.4kg and BMI (Wt/Ht) 24.0kg/m 2] and 1145 women (41.8y, 159cm, 58.6kg, 23.1 kg/m2). For the purpose of the study the following variables were considered: 1) individual characteristics such as age, sex, Ht and Wt; in addition body mass index was calculated as an index of fatness; 2) BIA variables: resistance (R), bioimpedance index (BI = height2/Z), reactance (Xc) and phase angle (PhA). BI (mean values: 58.0 vs 42.2cm2/ohm) and PhA (6.21 vs 5.76 degrees) were higher (P < 0.001) in men, while Z (611 vs 515 ohm) and Xc (61 vs 56ohm) were higher in women. After adjustment for age, Wt and BMI, these significant differences persisted (P< 0.001) for R, BI and PhA, but not for Xc (P = 0.78). The univariate analysis (simple correlation coefficient) revealed that BIA variables were significantly related to most of the individual characteristics considered. Indeed, R did not correlate with Ht in both sexes and Xc and PhA with Ht in women. Multiple regression analysis showed that age, Wt and BMI are all independent predictors of each of the BIA variables considered with the exception of BMI vs Xc in women and age vs PhA in men. The percentage of variability explained is much higher for BI (>50%) and lower for Xc and PhA (<20%), especially in women. In conclusion, BIA variables are significantly associated to the main characteristics of the subjects. These are of importance when considering for computation of percentiles for BIA variables or the graphical interpretation of BIA data.
P,73 Normal values for fat-free and fat mass in 1678 healthy adults measured by bioelectrical impedance analysis (BIA) U. G, Kyle 1, C. Pichard 1 and D. O. Slosman 2
1Clinical Nutrition, 2Nuclear Medicine, Geneva University Hospital, Geneva, Switzerland. Fat-free (FFM) and fat (FM) mass are useful in the evaluation of nutritional status in hospitalized patients if normal values for a healthy population are known. Reference values for FFM and FM as determined by BIA are not available. The purpose of this study was to determine the 25th and 75th percentiles for FFM and FM in a healthy population in Switzerland with the goal of developing FFM and FM nutritional assessment charts which permit rapid determination if hospitalized patients are below, within and above normal ranges for these nutritional parameters. Methods: Healthy men and women (n = 1678) without known pathology or medical treatment between the ages of 25 and 98, mean BMI of 21.5 _+ 2.7 kg/m 2 (women) and 22.0 _+2.0 kg/m 2 (men), were measured by BIA (BIO-Z2 ®, Spengler, Paris, France, 50 kHz, tetrapolar electrodes, right hand and foot). Height and body weight were measured with the precision of 0.5cm and 50g, respectively. Resistance values were determine twice. The subjects were grouped 25-44 years (n = 373 women, 589 men), 45-64 years (n = 167 women, 247 men), >65 years (n = 203 women, 99 men). The FFM and FM were estimated by the formulas by RJL. The percentiles of FFM and FM in kg were determined to develop assessment charts.
Results:
Age (yrs)
Women (25-75%ile) FFM (kg) FM (kg)
Men (25-75%ile) FFM (kg) FM (kg)
25~44 45-64 >65
40.4-46.5 41.1-47.3 38.6-45.4
56.8-65.0 8.8-13.8 5 5 . 8 - 6 3 . 1 9.4-16.1 50.6-59.7 11.9-25.1
11.3-16.4 12.8-20.3 15.7-26.8
P.75 Body density measurement by air displacement plethismography. A comparison with bioelectrical impedance and skinfold measurements F. Francini, R Spine//a, S. Mavrotheris, C. Serra, P. Di Lucchio and
Conclusions: A decrease in FFM and an increase in FM were observed with increasing age. Values below the 25th percentile help identify patients with abnormal lean and fat reserves. The use of assessment charts permits rapid comparison of FFM and FM reserves to normal values and aids in evaluation of nutritional status in hospitalized patients.
G. Bucciante Cattedra di Nutrizione Clinica, Universita di Padova, Padova, Italy. Introduction: The air displacement plethysmograph (ADP) is a new technique for the noninvasive determination of whole body volume and density. This technique provides many advantages over hydrodensitometry but up to date it has not been sufficiently compared with the other body composition methods. Subjects and methods: Sixty adult subjects (32 F, 28 M), varying in age (range 18-42yr) participated in this study. The subjects were in three groups according to BMI: normal weight (N.W., BMI: >18.5<25), under weight (U.W., BMI: <18.5) and over weight (O.W., BMI: >25). The body volume was measured through an air displacement plethysmograph (Bod Pod, Life Measurement Instruments, Concord, CA). The skinfold thickness was measured with a Harpenden caliper; the bioelectrical impedance was performed by an Akern 109 RJL The body density was calculated by the weight/volume ratio for the air displacement plethysmograph (DADP),by the Durrnin and Womersley equations for the skinfold
P,74 Relationships between BIA variables and individual characteristics M. Marra, V. BellizzP, B. R. Di Ioriot, V Terraccianot, G. Gaudianot and L. ScalfP Cattedra di Nutrizione Clinic& * Divisione di Nefrologia, §Dipartimento di Scienza degli Alimenti, Universit& degli Studi Federico II, Napoli; t Nefrologia, Ospedale di Lauria, Italy. Although bioimpedance analysis (BIA) is widely used for the estimation of body composition in healthy and ill subjects, the variability in the population of BIA measures is still poorly known. 49
Methods: The study population comprised 17 patients with cirrhosis without overt fluid retention (7M:10F) and 17 healthy volunteers (7M:10F). Hydrodensitometry, deuterium dilution and whole body mineral content obtained using DEXA (Hologic), were used in a four-component model to calculate hydration fraction and density of FFM. ~ Results: Values expressed as mean _+1SD:
thickness (DsT) and through total body water assessment for the bioelectrical impedance (DBIA). Data were analyzed using Student t-test. Results: The differences in body density values were significant between DADP (1.057g/cm 3) and DBIA (1.034g/cm 3, P < 0.0001). These differences between DADP and DBIA were also observed in the three groups: N.W. (DADP ---- 1065g/cmS; DBIA = 1.041, P < 0.0001), U.W. (DADe : 1.078 g/cmS; DBIA : 1.042 g/cm s, P < 0.005), O.W. (DADe = 1.033; DBIA : 1.019, P < 0.001); in the men (DADe = 1.066 g/cm3; DBIA = 1.042 g/cm 3, P < 0.005), O.W. (DAD P = 1.033; DBIA = 1.019, P < 0.001); in the men (DAD P = 1.066 g/cm3; DSIA = 1.041 g/cm s, P < 0.0001) and women groups (DADP = 1.046 g/cmS; DBIA = 1.024 g/cm 3, P < 0.0001). The differences between DADP and DST were significant only in the N.W. group (P< 0.01). Conclusions: The bioelectrical impedance overestimates the fat mass and this could explain the differences observed in this study.
Population
P. B. Soeters3 Depts of 1Dietetics, 2Clinical Chemistry and 3Surgery, University Hospital of Maastricht, Maastricht, The Netherlands.
Introduction: Bio-impedance spectroscopy (BIS) is a simple, non-invasive, portable method to measure body composition. The method is very attractive for routine use in clinical settings. However, validation studies described often yield different results. This can partly be explained by the different approaches to transform measured resistance values into body compartments. Aim: To compare the linear regression (LR) method with the Hanai mixture theory (HM) for the conversion of measured resistance values into body fluid compartments measured by dilution methods. Methods: In 91 persons with different degree of obesity (BMI: mean 39, range 23-62) total body water (TBW) and extracellular water (ECW) were measured by deuterium and NaBr dilution methods respectively. BIS was measured with a Xitron 4000B model instrument uysing 48 frequencies ranging from 5 to 500 kHz. Fitting the data to the Cole-Cole model yielded resistance of the ECW (Recw) and TBW (Rtbw). Data of half the group were used for LR based on L2/R to predicts TBW and ECW and to calculate the constants used in the HM (kecw and kp). TBW and ECW in the other half of the group were calculated with the prediction equations based on LR and on HM using the new constants. The accuracy of both methods were compared. Results: TBW and ECW according to the dilution methods were 40.05 I (sd 6.8) and 19.441 (sd 3.09). Comparison with LR and HM methods:
TBW LR TBW HM ECW LR ECW HM
SEE
Correlation
38.98 39.72 18.32 18.64
1.07 0.33 1.13 0.80
2.38 2.14 1.77 1.73
0.941 0.950 0.863 0.858
(5.80)* (6.67) (3.52)* (3.34)*
Hydration fraction FFM
Density of FFM (kg/I)
0.764_+0.028 0.735_+0.019 <0.0025
1.097-+0.009 1.102_+0.007 <0.025
The hydration fraction and density of FFM were inversely correlated in both patients (r = -94, P < 0.0001) and volunteers (r = -0.73, P < 0.005). The mean differences in hydration fraction (3.9%) and density (0.5%) indicate over- and underestimation of FFM by 4.5% and 1.1% respectively, if standard figures are used. ~ Conclusion; In patients with cirrhosis both hydrodensitometry and water dilution produce errors in the estimation of FFM even in the absence of overt fluid retention which preclude their use as single body composition techniques in this patient population.
P. L. M. C o x - R e i j v e n ~, B. van KreeP, M. P. Geerlings ~ and
Mean difference with dilution
BMI (kg.m-2)
Patients (n = 17) 47.6_+7.4 26.1 _+4.3 Volunteers(n=17) 47.1 _+8.3 25.6_+3.7 'P' value ns ns
P.76 Validation of bio-impedance spectroscopy, comparison of regression analysis with the Hanai mixture theory
Mean (sd)
Age (yr)
Reference: 1. Fuller. Clin Sci 1992; 82: 687.
P.78 Validity of body fat assessment in patients with cirrhosis using anthropometry and DEXA A. M. M a d d e n 1, N. J. Fuller2, G. Jennings2, M. Ella2 and M. Y. Morgan1
I Dept of Medicine, Royal Free Hospital, London and 2MRC Dunn Clinical Nutrition Centre, Cambridge, UK. Body composition data are required in patients with cirrhosis to define the pattern of tissue loss, quantify metabolically active tissue and monitor the effects of nutrition intervention. Estimates of body fat are frequently derived using anthropometry but the use of DEXA for this purpose is becoming more widespread. Aim: To investigate the use of anthropometry and DEXA for estimating body fat in patients with cirrhosis in comparison with a reference fourcomponent model. 1 Methods: Ten men and 10 women (mean _+ 1SD age 48.6 + 8.1 yr; BMI 26.1 _+4.5), noine of whom had oedema or ascites, were assessed using: i) Four-component model (4CM) 1 based on combined data from hydrodensitometry, deuterium dilution and whole body mineral obtained using DEXA; ii) anthropometry using skinfold thicknesses (AN); 2 iii) DEXA (Hologic QDR-1000/W scanner with Enhanced Whole Body V5.61 programme). Results: Values expressed as mean _+1 SD:
* Sign. different from dilution (P < 0.005), paired t-test. The constant Kecw for calculating ECW was sign. correlated with BMI (r = -0.352, P < 0.001 ). Conclusion: The Hanai mixture theory gives better results than linear regression analysis and has the advantage of not being population specific. However, the 'constant' used for calculating ECW is not constant but decreases with an increasing degree of obesity. Correction of Kecw for degree of obesity will improve accuracy of the results.
Population
% Body fat 4CM
Difference in % body fat 4CM-AN 4CM-DEXA
Men (n = 10) Women (n = 10)
29.7 _+9.8 35.6_+7.2
9.6 _+5.4** 3.3_+5.7
5.9 _+2.4** 4.0_+3.4*
Difference from 4CM: * P < 0.005; ** P < 0.0005 (paired t-tests).
Conclusion: In this patient population both AN and DEXA underestimated body fat measurements determined using the 4CM, particularly in men. The use of DEXA as a reference method is not recommended in this patient population.
Reference: 1. Fuller. Clin Sci 1992; 82: 687. 2. Dumin. BrJ Nutr 1974; 32: 77.
P.77 Alterations in the density and hydration fraction of fat free mass in patients with cirrhosis without overt fluid retention
P.79 Determination of body water distribution in ascitic subjects
G. Jennings 2, M. Ella2 and M. Y. Morgan1 ~Dept of Medicine, Royal Free Hospital, London and 2MRC Dunn Clinical Nutrition Centre, Cambridge, UK.
L. J. C. Bluck, A. M. Madden*, M. Y. Morgan*, M. Ella and G. Jennings Dunn Nutrition Unit, Cambridge, UK, *Dept of Medicine, Royal Free Hospital, London, UK.
Hydrodensitometry and water dilution techniques are classic methods of assessing body composition, but they rely on assumptions about the density (1.10 kg/I) and hydration fraction (0.73) of fat free mass (FFM). Aim: To test the hypothesis that in cirrhotic patients without overt fluid retention there is an increase in hydration fraction and an associated decrease in the density of FFM.
Determination of body water in ascitic subjects using the standard deuterium dilution technique has produced variable results because of the variable time taken for label to equilibrate with the ascitic fluid and any other excess fluid in a secondary pool. In addition, the technique gives no information about the mass of the secondary pool. 50