- Email: [email protected]
Reference values of fat-free and fat masses by bioelectrical impedance analysis in 3393 healthy subjects
APPLIED NUTRITIONAL INVESTIGATION
Reference Values of Fat-Free and Fat Masses by Bioelectrical Impedance Analysis in 3393 Healthy Subjects Claude Pichard, MD, PhD, Ursula G. Kyle, MS, RD, David Bracco, MD, Daniel O. Slosman, MD, Alfredo Morabia, MD, PhD, and Yves Schutz, PhD From the Divisions of Clinical Nutrition, Nuclear Medicine, and Epidemiology, Geneva University Hospital, Geneva; and the Institute of Physiology, Lausanne, Switzerland Determination of fat-free mass (FFM) and fat mass (FM) is of considerable interest in the evaluation of nutritional status. In recent years, bioelectrical impedance analysis (BIA) has emerged as a simple, reproducible method used for the evaluation of FFM and FM, but the lack of reference values reduces its utility to evaluate nutritional status. The aim of this study was to determine reference values for FFM, FM, and %FM by BIA in a white population of healthy subjects, to observe the changes in these values with age, and to develop percentile distributions for these parameters. Whole-body resistance of 1838 healthy white men and 1555 women, aged 15– 64 y, was determined by using four skin electrodes on the right hand and foot. FFM and FM were calculated according to formulas validated for the subject groups and analyzed for age decades. This is the first study to present BIA-determined age- and sex-specific percentiles for FFM, FM, and %FM for healthy subjects, aged 15– 64 y. Mean FM and %FM increased progressively in men and after age 45 y in women. The results suggest that any weight gain noted with age is due to a gain in FM. In conclusion, the data presented as percentiles can serve as reference to evaluate the normality of body composition of healthy and ill subject groups at a given age. Nutrition 2000;16:245–254. ©Elsevier Science Inc. 2000 Key words: bioelectrical impedance analysis, fat-free mass, body fat, fat-free mass measured by bioelectrical impedance analysis, body composition, reference standard, gender
INTRODUCTION Determination of fat-free mass (FFM) and fat mass (FM) is of considerable interest in the evaluation of nutritional status. Both over- and undernutrition contribute to increased mortality and morbidity. The most common estimate of excess or deficit in body weight is the body mass index (BMI). However, BMI has been shown to be an imprecise measurement of fatness.1,2 Simple measurements for evaluating body composition such as skinfold measurements are easy to perform but lack accuracy and reproducibility.1 Other methods require subject cooperation (underwater weighing) or sophisticated equipment and skilled technicians (tracer dilution, dual-energy x-ray absorptiometry (DXA), and neutron activation analysis). To date, the only measures used to grade body composition in very large studies (such as the Study by Metropolitan Life Insurance) have been relative weight, weight for height, or BMI. In recent years, bioelectrical impedance analysis (BIA) has emerged as a simple, reproducible method that can be used for the evaluation of FFM and FM. Because of its low cost and ease and rapidity of use, one of the greatest potential uses of BIA is in studies of large populations. BIA represents a substantial improvement in the epidemiologic method of body composition compared with BMI1; it seems to be valid for subjects without large disturbances in fluid distribution and provides a more reliable measure-
Correspondence to: Claude Pichard, MD, PhD, Clinical Nutrition and Diet-therapy, Geneva University Hospital, 1211 Geneva, Switzerland. Email: [email protected] Date accepted: Jan. 19, 2000. Nutrition 16:245–254, 2000 ©Elsevier Science Inc., 2000. Printed in the United States. All rights reserved.
ment of body composition with respect to FFM and FM than do simpler methods such as skinfolds and height and weight.3,4 The aims of this study were to determine reference values for FFM, FM, and %FM by BIA in a large white (Western European) population of healthy subjects in Switzerland, to observe differences in these values in age groups of 10 y between ages 15 and 64 y, and to develop percentile distributions for these parameters. This is the first study that examines FFM and FM by BIA in an adult population across the age spectrum and can serve as a baseline to monitor future trends in body composition and changes in FFM, FM, and %FM. Moreover, age- and sex-specific percentile distributions, not available until the present time, will permit more meaningful interpretations of FFM and FM. The data presented as percentiles can serve as references to evaluate the normality of body composition of healthy or ill groups at a given age.
SUBJECTS AND METHODS Subjects To determine reference values for a healthy subject population, 3393 healthy adults (1838 men and 1555 women), aged 15– 64 y, were recruited by offering free BIA on an exhibition stand at trade fairs and fun runs and among public administration staff. This method allowed us to include a wide representation of people with different levels of physical activity. The anthropometric data and number of healthy subjects per age group is shown in Table I. All subjects were ambulatory, with no known pathologies or physical handicaps. Subjects were questioned on use of medications and reasons for visits to a physician in the last 6 mo to eliminate subjects with acute or chronic diseases. Due to the heterogeneity of 0899-9007/00/$20.00 PII S0899-9007(00)00256-2
246
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS TABLE I. ANTHROPOMETRIC AND BIOIMPEDANCE CHARACTERISTICS (MEAN ⫾ SD) OF A HEALTHY WHITE POPULATION
Age (y) Men n Height (cm) Weight (kg) BMI (kg/m2) IBW (%) Resistance (⍀) Reactance (⍀) Women n Height (cm) Weight (kg) BMI (kg/m2) IBW (%) Resistance (⍀) Reactance (⍀)
15–64*
15–24
25–34
35–44
45–54
55–64
1838 176.3 ⫾ 6.7 (155–199) 73.2 ⫾ 9.2 (48.1–117.6) 23.5 ⫾ 2.7 (17.1–40.1) 108.4 ⫾ 13.0 (74–172) 475 ⫾ 48 (334–743) 60.7 ⫾ 9.9 (28.8–113)
368 177.5 ⫾ 6.5 69.7 ⫾ 8.2 22.1 ⫾ 2.1 101.9 ⫾ 10.3 483 ⫾ 53 63.3 ⫾ 9.5
503 177.6 ⫾ 6.6 73.3 ⫾ 88 23.2 ⫾ 2.3 106.2 ⫾ 12.1 475 ⫾ 47 62.8 ⫾ 10.5
489 176.6 ⫾ 6.5 74.7 ⫾ 9.5 24.0 ⫾ 2.8 110.1 ⫾ 13.1 472 ⫾ 50.2 60.0 ⫾ 10.1
306 174.1 ⫾ 6.3 73.9 ⫾ 9.2 24.4 ⫾ 2.6 112.7 ⫾ 11.5 473 ⫾ 42 58.9 ⫾ 8.5
172 173.4 ⫾ 6.5 75.3 ⫾ 9.5 25.0 ⫾ 3.0 116.6 ⫾ 14.0 472 ⫾ 46 55.3 ⫾ 11.1
1555 164.5 ⫾ 6.3 (147–186) 58.9 ⫾ 8.2 (39.1–99.8) 21.8 ⫾ 2.9 (16.8–37.7) 105.1 ⫾ 13.7 (73–165) 580 ⫾ 61 (414–837) 67.4 ⫾ 13.8 (32–132)
407 166.5 ⫾ 6.2 57.7 ⫾ 7.7 20.8 ⫾ 2.4 101.0 ⫾ 12.1 601 ⫾ 61 72.0 ⫾ 14.2
425 165.4 ⫾ 6.1 58.3 ⫾ 7.8 21.3 ⫾ 2.5 101.8 ⫾ 13.1 582 ⫾ 61 66.5 ⫾ 13.8
344 163.6 ⫾ 5.9 58.7 ⫾ 8.1 21.9 ⫾ 2.7 105.8 ⫾ 13.0 570 ⫾ 55 67.7 ⫾ 15.1
247 163.2 ⫾ 5.9 60.0 ⫾ 8.4 22.5 ⫾ 2.8 109.3 ⫾ 13.0 562 ⫾ 58 64.4 ⫾ 13.6
132 160.5 ⫾ 5.9 63.5 ⫾ 9.4 24.7 ⫾ 3.6 114.2 ⫾ 15.1 564 ⫾ 65 61.7 ⫾ 13.7
* Ranges are given in parentheses. BMI, body mass index; IBW, ideal body weight (Metropolitan Life Insurance, 1983).
the population (about one-third of population is of non-Swiss nationality), the sample is more representative of a European than strictly Swiss population. Subjects older than 65 y were excluded because of insufficient validation of BIA formulas in elderly subjects. The study protocol complied with the requirements of the Geneva University Hospital Ethics Rules. Anthropometric Measurements and Bioelectrical Impedance Analysis Body height was measured to the nearest 0.5 cm and body weight to the nearest 0.1 kg on a balance beam scale (Seca Corp. Mechanical Scale, Berlin, Germany). Resistance and reactance were measured by a BIA generator and used to mathematically derive FFM and FM, as previously described,5–7 by using the formula V ⫽ ⫻ ht2/R, in which conductive volume (V) is assumed to represent FFM, is the specific resistivity of the conductor, height (ht) is taken as the length of the conductor, and body resistance (R) is measured with four surface electrodes placed on the right wrist and ankle. Briefly, an electrical current of 50 kHz and 0.8 mA was produced by a generator (Bio-Z2, Spengler, Paris, France) and applied to the skin with adhesive electrodes (3M Red Dot, 3M Health Care, Borken, Germany) with the subject in the decubitus dorsalis.8 The skin was cleaned with 70% alcohol. The technical range of the Bio-Z2 generator is: impedance, 100 to 900 ⍀; phase angle, 1 to 18°; and an accuracy of ⫾3 ⍀ and ⫾0.2°, respectively. Short- and long-term reproducibility of resistance measurements indicate coefficients of variance of 1.8 –2.9%.9,10 Several BIA instruments were used in this study to permit inclusion of a large number of subjects at one time. The Bio-Z2 generators were cross-validated at 50 kHz against the RJL-109 and 101 analyzers (RJL Systems Inc., Clinton Township, MI, USA) and against the Xitron analyzer (Xitron Technologies, Inc., San Diego, CA, USA). Limit of tolerance between instruments was ⫾5 ⍀ at 50 kHz using a calibration jig and in vivo measurements. All data collectors had standard training and were provided verbal and written information to standardize the procedure and minimize errors due to multiple data collectors. All measurements were made 2–3 h after meal intake, and measurements of participants in fun runs were made before the run
to avoid changes in hydration, skin temperature, electrolyte concentration, and glycogen stores.11 Validation of BIA Formulas Body composition was calculated with formulas shown in Table II. Validation of FFM against DXA (Hologic QDR-4500, Whole Body software version, Hologic Inc., Waltham, MA, USA) in subgroup of 80 non-obese men and 69 non-obese women resulted in a correlation of r ⫽ 0.97 and 0.96, standard error of the estimate (SEE) of 2.0 and 1.6 kg, coefficient of variation (CV) of 3.7% and 4.1%, and mean difference of ⫺0.1 ⫾ 2.0 kg and ⫺0.4 ⫾ 1.6 kg in men and women, respectively (unpublished data), confirming the appropriateness of the BIA formula in the
TABLE II. FORMULAS TO CALCULATE FAT-FREE MASS USING BIOELECTRICAL IMPEDANCE ANALYSIS* Body density† %Fat mass FFM Overweight FFM‡ Women %Fat mass†
FFM Overweight FFM‡
1.1554 ⫺ (0.0841 ⫻ weight ⫻ resistance/[height2]) (4.95/body density ⫺ 4.5) ⫻ 100 Weight ⫺ fat mass BMI ⬎26 kg/m2 14.5244 ⫹ 0.000886 ⫻ (height2) ⫺ 0.02299 ⫻ R ⫹ 0.42688 ⫻ weight ⫺ 0.07002 ⫻ age (1 ⫺ [0.3981 ⫻ height2/resistance ⫹ 0.3066 ⫻ weight ⫹ 0.095299 ⫻ {height ⫺ 100} ⫹ 0.7414]/weight) ⫻ 100 Weight ⫺ fat mass BMI ⬎ 30 kg/m2 9.3794 ⫹ 0.000912 ⫻ (height2) ⫺ 0.01466 ⫻ R ⫹ 0.2999 ⫻ weight ⫺ 0.07012 ⫻ age
* Different formulas were used for overweight subjects. † Manufacturer’s formula (RJL Systems Inc.). ‡ From Segal et al.14 BMI, body mass index; FFM, fat-free mass.
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
247
TABLE III. PERCENTILES FOR FAT-FREE MASS (kg) BY BIOELECTRICAL IMPEDANCE ANALYSIS FOR WHITE ADULTS Percentile Age (y) group Men Total 15–24 25–34 35–44 45–54 55–64 Women Total 15–24 25–34 35–44 45–54 55–64
n
Mean ⫾ SD
5th
10th
25th
50th
75th
90th
95th
1838 368 503 489 306 172
60.3 ⫾ 6.2 59.5 ⫾ 6.7 61.1 ⫾ 6.3* 61.0 ⫾ 6.2 59.3 ⫾ 5.4* 59.1 ⫾ 5.6
50.3 48.2 50.7 51.6 50.2 51.0
52.5 50.9 52.9 53.3 52.5 52.8
55.8 55.0 56.6 56.5 55.8 55.0
60.4 59.7 61.6 60.9 59.0 59.0
64.4 63.9 65.0 65.3 62.8 62.6
68.1 68.0 68.7 69.3 66.3 66.6
70.7 70.4 71.2 71.6 69.4 69.3
1555 407 425 344 247 132
43.7 ⫾ 4.5 43.3 ⫾ 4.7 43.7 ⫾ 4.4 43.6 ⫾ 4.5 44.1 ⫾ 4.4 44.0 ⫾ 4.5
36.6 35.8 36.6 37.0 36.6 37.0
37.8 37.3 38.1 37.8 37.8 38.3
40.5 40.2 40.4 40.5 41.6 40.3
43.6 42.9 43.7 43.7 44.3 43.8
46.6 46.2 46.7 46.6 46.6 47.1
49.5 49.6 49.0 49.8 49.5 49.6
51.4 51.4 51.4 51.1 51.6 51.5
* P ⬍ 0.001 versus preceeding age group, analysis of variance.
subjects of this study. Results of validation are shown in Annex 1 (anthropometric data). Simple regressions and Bland-Altman analysis of non-obese subjects are shown in Annexes 2 (men) and 3 (women). Several studies12–14 have shown that BIA formulas developed for normal-weight subjects are not adapted to overweight subjects. Therefore, we used formulas by Segal et al.14 for overweight subjects (Table II) to calculate FFM in men with BMI greater than 26 kg/m2 and in women with BMI greater than 30 kg/m2. These cutoff values were obtained from our validation of BIA against DXA (unpublished data). The calculations by Segal et al. were used in 291 (15.8%) men and 53 (3.4%) women who fell in the overweight/obese category. Validation of FFM against DXA in 58 obese men and 20 obese women resulted in a correlation of r ⫽ 0.95 and 0.82, SEE of 2.2 and 3.0 kg, CV of 3.6% and 5.1%, and mean difference of 0.4 ⫾ 2.3 kg and 0.9 ⫾ 2.4 kg in men and women, respectively (unpublished data), confirming the appropriateness of the BIA formula in the obese subjects. Simple regressions and Bland-Altman analysis of validation in obese subjects are shown in Annexes 4 (men) and 5 (women).
Statistics The statistical analysis program StatView, version 4.1 (Abacus Concepts, Berkeley, CA, USA) was used for statistical analysis. The results are expressed as mean ⫾ standard deviation. Age- and sex-specific percentile distributions were calculated for each of the following parameters: FFM, FM, and %FM. To compare our results with those of other studies, the data were stratified by steps of 10 y as reported for BMI and anthropometric data in the NHANES study15,16 and a Canadian study.17 The ith percentile (Pi) was the value at or below which there was i% of the sample. For example, the 50th percentile (P50) was the value at or below which there were 50% of the observations for a given variable. Given a total of n ordered values for each parameter ( x 1 , x 2 , x 3 , . . . , x n ), Pi in any of the calculated distribution was computed as follows with the Statview 4.1 statistical program: Pi ⫽ (1 ⫺ A) ( x b ) ⫹ ( A) ( x b ⫹ 1 ). The differences between age groups were analyzed by analysis of variance with Fisher’s protected least-significant-difference comparison.
RESULTS Table I shows the anthropometric and bioimpedance characteristics of the men and women with no known pathologies. Age- and sex-specific percentile distributions for FFM, FM, and %FM of these adults are presented in Tables III–V and in Figures 1–2. Body Mass Index The mean BMI increased from 23.2 kg/m2 in men and 21.3 kg/m2 in women aged 25–34 y to 25.0 kg/m2 in men and 24.7 kg/m2 in women aged 55– 64 y (Table I). Fat-Free Mass The mean FFM (Table III) was 60.3 ⫾ 6.2 kg for white men between 15 and 64 y. Of these men, 90% had an FFM between 52.5 and 68.1 kg. The mean FFM was greatest for men aged 25– 44 y. Mean FFM was significantly higher in men aged 25–34 y than in men aged 15–24 y and significantly lower in men aged 45–54 y than in men aged 35– 44 y. The percentiles developed for men between 25 and 44 y generally exceeded comparable percentiles calculated for men of other age groups. A diagram of these trends is presented in Figure 1. For women, the mean FFM was 43.7 ⫾ 4.5 kg. Ninety percent of women had an FFM between 37.8 and 49.5 kg. The FFM increased slightly but non-significantly from 43.3 ⫾ 4.7 kg for women aged 15–24 y to 44.1 ⫾ 4.4 kg for women aged 45–54 y. These age-related trends were also apparent in the percentile distributions. Figure 1 displays these trends in graphic form. Fat Mass and Percentage of Fat Mass The mean FM (Table IV) for white men between the ages of 15 and 64 y was 12.9 ⫾ 5.7 kg. Of these men, 90% had an FM between 7.4 and 22.1 kg. FM of men showed differences related to aging. The FM increase was significant for all age groups. The mean FM was 11.1% greater in those 35– 44 y old and 32.8% greater in those 55– 64 y old than in those 25–34 y old. The mean %FM (Table V) for white men was 17.3 ⫾ 5.8%. Of these men, 90% had a %FM between 11.1 and 27.1%. The parallel differences
248
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS TABLE IV. PERCENTILES FOR FAT MASS (kg) BY BIOELECTRICAL IMPEDANCE ANALYSIS FOR WHITE ADULTS Percentile
Age (y) group Men Total 15–24 25–34 35–44 45–54 55–64 Women Total 15–24 25–34 35–44 45–54 55–64
n
Mean ⫾ SD
5th
10th
25th
50th
75th
90th
95th
1838 368 503 489 306 172
12.9 ⫾ 5.7 10.2 ⫾ 3.7 12.2 ⫾ 4.7* 13.6 ⫾ 5.8* 14.5 ⫾ 6.0† 16.2 ⫾ 7.2*
6.6 6.1 6.6 6.8 7.4 7.0
7.4 6.8 7.4 7.7 8.3 8.1
9.0 7.9 8.9 9.6 9.9 10.6
11.3 9.4 11.2 12.1 13.1 14.3
15.1 11.4 14.3 16.5 18.3 22.7
22.1 14.5 19.3 22.2 23.5 26.6
24.5 17.1 22.3 24.8 26.2 27.5
1555 407 425 344 247 132
15.2 ⫾ 5.0 14.3 ⫾ 4.0 14.6 ⫾ 4.6 15.1 ⫾ 4.9 15.8 ⫾ 5.3 19.5 ⫾ 6.5*
9.0 8.8 8.9 9.0 8.8 11.2
9.9 9.9 9.6 9.9 9.7 12.8
11.9 11.7 11.6 12.0 12.3 15.1
14.4 13.8 14.0 14.2 15.6 19.1
17.4 16.4 16.7 17.1 19.1 21.9
21.0 19.2 19.4 21.2 21.7 26.0
23.5 21.0 21.7 22.4 23.2 31.7
* P ⬍ 0.001 versus preceeding age group, analysis of variance. † P ⬍ 0.05 versus preceeding age group, analysis of variance.
between age groups were not consistently present in the percentile distribution. Greater dispersion was noted in the upper percentiles and older subjects. The mean FM for women (Table IV) between the ages of 15 and 64 y was 15.2 ⫾ 5.0 kg. Of these women, 90% had an FM between 9.9 and 21.0 kg. Marked differences were observed between the FM means determined for the group older than 55 y. The largest mean, observed in those aged 55– 64 y, exceeded the smallest mean, noted in those aged 15–24 y, by 36.4%. The mean %FM for white women was 25.3% (Table V). Of these women, 90% had an FM between 19.2 and 31.6%. A significant increase was only noted in %FM in women older than 55 y. Percentile distributions were parallel in women aged 15–54 y, but showed marked differences between age groups in those older than 55 y,
with increases noted in all percentile distributions. Figure 2 shows trends in FM, and Figure 3 shows trends in %FM for men and women.
Gender Differences in Fat-Free Mass and Fat Mass As expected, gender differences were noted in FFM. The mean FFM was 38% greater in men than in women and was slightly higher in 25–34-y-old subjects than in those older than 45 y. The age distribution showed that the mean FFM remained in a narrow range of 59.1– 61.0 kg for men and 43.3– 44.1 kg for women between ages 15 and 64 y. The higher FFM in men and relatively stable FFM throughout the adult age range is confirmed by a mean
TABLE V. PERCENTILES FOR FAT MASS (%) BY BIOELECTRICAL IMPEDANCE ANALYSIS FOR WHITE ADULTS Percentile Age (y) group Men Total 15–24 25–34 35–44 45–54 55–64 Women Total 15–24 25–34 35–44 45–54 55–64
n
Mean ⫾ SD
5th
10th
25th
50th
75th
90th
95th
1838 368 503 489 306 172
17.3 ⫾ 5.8 14.5 ⫾ 4.3 16.3 ⫾ 4.9* 17.8 ⫾ 5.8* 19.2 ⫾ 6.0* 20.9 ⫾ 7.2*
10.1 9.4 10.1 10.3 11.3 10.3
11.1 10.4 10.9 11.7 12.4 11.7
13.1 11.9 12.9 13.5 14.7 15.2
15.8 13.7 15.3 16.4 17.7 19.1
20.0 16.1 18.8 21.4 24.5 28.4
27.1 19.3 24.0 26.9 28.5 30.7
28.8 22.5 26.2 28.7 29.7 32.0
1555 407 425 344 247 132
25.4 ⫾ 5.1 24.5 ⫾ 4.2 24.6 ⫾ 4.7 25.2 ⫾ 5.1 25.9 ⫾ 5.5 30.1 ⫾ 5.8*
17.6 18.1 17.4 17.7 16.9 22.3
19.2 19.6 18.8 19.1 18.2 24.3
22.0 21.7 21.5 21.9 22.4 26.1
25.1 24.4 24.4 24.8 26.4 30.0
28.3 27.1 27.2 28.2 29.5 32.8
31.6 29.8 30.2 31.6 31.9 36.0
33.4 31.3 31.9 32.9 33.6 42.6
* P ⬍ 0.001 versus the preceeding age group, analysis of variance.
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
249
FIG. 1. Apparent changes in fat-free mass of white adults with advancing age.
FFM index (kg FFM/ht2; data not shown) of 19.3 kg/cm2 in 25–34-y-old men and 19.6 kg/cm2 in 35– 64-y-old men and of 16.0, 16.3, 16.5, and 17.0 kg/m2 in 25–34, 35– 44, 45–54, and 55– 64-y-old women, respectively. Gender differences were also noted in FM and %FM. The higher FM in women is confirmed by higher mean FM indexes (kg FM/ht2) of 5.2, 5.3, 5.6, 5.9, and 7.6 kg/m2 in women and 3.2, 3.8, 4.3, 4.8, and 5.4 kg/m2 in men aged 15–24, 25–34, 35– 44, 45–54, and 55– 64 y, respectively. FM indexes increased significantly (P ⬍ 0.01) in all age groups except in women 25–34 y versus those 15–24 y and in women 55– 64 y versus those 45–54 y. A larger distribution in %FM was noted in women. Mean %FM was
FIG. 2. Apparent changes in fat mass of white adults with advancing age.
significantly greater in women than in men. The 95th percentile for %FM was 42.6% for women versus 32.0% for men, suggesting a greater increase with age in %FM in women than in men. Furthermore, larger FM and %FM in older adults showed that any weight gain was explained by FM gain in both sexes.
DISCUSSION Few studies have reported on FFM and %FM in large population samples,1,18,19 and none have reported on sex- and age-specific percentiles for measured FFM, FM, and %FM. Most studies that
250
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
FIG. 3. Apparent changes in percentage of fat mass in white adults with advancing age.
have reported on FM are based on skinfold measurements and other indirect estimations such as waist circumference and waistto-hip ratios.17,20 BIA is being recognized as an easy, portable method that allows estimation of body composition compartments. Despite concerns about the validity of BIA when compared with sophisticated reference techniques, it remains the only technique that is inexpensive and applicable to large numbers of subjects.1,4 As far as we know, ours is the first study to examine FFM and FM by BIA in a large, white, healthy population and evaluate body composition parameters across the age spectrum from 15 to 64 y. The percentiles derived from these results can serve as references for clinical evaluation of body composition in healthy and ill subjects. Validation of BIA Formulas Validity of FFM and FM obtained by BIA depends on the use of an equation appropriate for the subset of the population.1,21,22 Different BIA formulas were used in men and women and in overweight and normal-weight subjects. These formulas were validated (see SUBJECTS AND METHODS and Annexes 1 through 5) against DXA. The validation confirmed the appropriateness of the BIA formula in healthy subjects participating in the present study, including overweight subjects. Body Mass Index The BMI in this study serves to compare the subjects with those in other epidemiologic studies. Because BMI is a rough indicator of fatness, studies including subjects with higher BMI, such as the Framingham studies in the United States,1 would be expected to report higher %FM than that reported in this study. The BMI of our study was slightly lower than the BMI of 25.3 kg/m2 in men and 23.0 kg/m2 in women reported a randomly sampled Swiss population (ages 40 –59 y) evaluated as part of a community-based surveillance program (the Euralim Study, a European public-health study of eating habits and cardiovascular disease and cancer risk factors).23 The difference is due to our population including younger subjects (60% of subjects ⬍40 y) and proportionally fewer obese subjects. Nineteen percent of subjects had a BMI greater than 25 kg/m2 compared with 29.2%
reported in the Fourth Nutrition Report in Switzerland (n ⫽ 510). 24 However, this difference may be due to an underrepresentation of older persons (8.5% versus 29% ⬎55 y) in our study and the exclusion of obese subjects with secondary pathologies. The BMI in this study is comparable to a mean BMI of 24.6 ⫾ 3.6 kg/m2 in men and 25.3 kg/m2 in women in Finland25 and a BMI of 21.5 kg/m2 at age 20 y and 25.4 kg/m2 at age 60 y in men and a BMI of 20.6 kg/m2 at age 20 y and 24.1 kg/m2 at age 60 y in women in France.26 Thus, mean BMI in Swiss subjects is similar to that in other European populations.27 The BMI was lower than BMIs of 28.6 ⫾ 4.4 kg/m2 in men aged 30 –39 y and 29.0 ⫾ 3.5 kg/m2 in men aged 60 – 69 y and BMIs of 27.2 kg/m2 in women aged 30 to 39 y and 28.0 ⫾ 5.7 kg/m2 in women aged 60 – 69 y reported by the Framingham Offspring Study in the United States.1 Important differences in BMI are noted in men younger than 54 y and in women younger than 64 y when compared with the large, nationally representative North American white sample,28 which had BMIs 4 –7 kg/m2 higher at the 90th percentile than the white subjects in this study.27 Recent studies have suggested significant increases in mean BMI in North Americans in the last 20 y.29 Such an increase is not apparent in Switzerland,24 although representative data of the whole population are lacking. Fat-Free Mass No study has reported changes of FFM in large population samples over several decades. Progressive decline in FFM with advancing age reported in a longitudinal study30 could not be confirmed in the present cross-sectional study. Bartlett et al.19 noted a decline beginning at age 60 y in men. Mean FFM decreased slightly but was statistically significant in men aged 45–54 y versus men aged 35– 44 y in our study, but this mean decrease in FFM with age was not significant when corrected for the shorter stature of older men. The earlier slight decrease in FFM in women, from 51–55 y, reported by Bartlett et al.19 was not confirmed in our study. In women, FFM normalized for height shows a slight increase in FFM until age 64 y, probably related to heavier subjects having greater FFM. Bartlett et al.19 determined FFM and FM by underwater weighing in different age groups consisting of approximately 1000 healthy subjects. They found similar FFM (60.0 – 62.2 kg) in men
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS aged 20 – 60 y and higher FFM (45.0 – 47.6 kg versus 43.3– 44.1 kg in the present study) in women who had higher BMIs, thus confirming FFM as being similar in different populations as measured by different body composition methods. The Fels Longitudinal Study31 reported 59.4 kg of FFM (by DXA) in men aged 20 –29 y and 60.2 kg in men older than 60 y and 42.2 kg of FFM in women aged 20 –29 y, 41.1 kg in women aged 50 –59 y, and 38.3 kg in women older than 60 y; women older than 60 y had a lower BMI than did women aged 50 to 59 y, suggesting that FFM remains stable if body mass does not decrease. A drop in FFM usually occurs after age 65 y, an age category not included in this study because of lack of adequate validation of BIA formulas in elderly subjects. The results of this study suggest that the mean FFM remains relatively stable in healthy adults throughout adulthood until approximately 60-y-old. The average body mass was progressively higher in each category of age (about ⫹5 kg in 40 y).27 Because body mass gain is accompanied by an increase in both FM and FFM, this increase will offset the drop in FFM with advancing age expected at a normal body weight. Fat Mass and Percentage of Fat Mass The results of this study showed that body weight and BMI increased until age 64 y and that the increase was predominantly due to an increase in the FM. Numerous studies have reported increases in subscapular and triceps skinfolds, which represent FM, with advancing age.25,32 This study agrees with previous reports and shows a significant increase in FM and %FM in men in all age ranges, which is parallel to increases in weight and a trend of an increase in FM and %FM in women, but which reached statistical significance only in women 55– 64-y-old as opposed to women 45–54-y-old. Because the data are cross-sectional rather than longitudinal, trends in body composition parameters observed with advancing age may have represented differences between successive generations as well as physiologic alterations with aging. Roubenoff et al.1 found higher %FM (29.9 ⫾ 6.1% in men and 37.7 ⫾ 7.8% in women) in subjects with higher BMIs participating in the Framingham studies and noted a mean increase in %FM with advancing age in both sexes, peaking in the 60s and 50s for American women and men, respectively, which is inconsistent with our results of an increase in %FM in men until age 64 y. Biasoli et al.33 reported similar %FMs by BIA in a small number of subjects with higher BMIs. The relative lower FM may be related to using the same formula for overweight subjects; thus, FM may have been underestimated in their more obese subjects. Higher %FM by hydrodensitometry in men and women with higher weights and BMIs than those in our study have been reported by Bartlett et al.34 The difference may also have been due to inclusion in our study of some subjects with relatively high physical activity levels and thus lower FM. Keharias et al.35 also found higher mean %FM (ranging from 14.9 –25.7% in men and from 28.0 –35.4% in women) by neutron activation analysis in subjects with higher BMIs. Similarly, the Fels Longitudinal Study31 reported 19.5%FM in men aged 20 –29 y with a mean BMI of 23.1 kg/m2 and 31.2%FM in men older than 60 y with a BMI of 29.8 kg/m2 and 34%FM in women aged 20 –29 y with a BMI of 23.6 kg/m2 and 39.4%FM in women older than 60 y with a mean BMI of 24.4 kg/m2. Comparisons with other studies in which height, weight, and age are significantly different are difficult to make. Guo et al.36 reported higher FM and %FM by underwater weighing in men (n ⫽ 102) and women (n ⫽ 109) aged 40 – 66 y who had higher BMIs than those reported in our study and found a significant decrease in FFM and increases in total FM, %FM, body weight, and BMI with advancing age in their longitudinal study (mean follow-up of 9 y). Thus, changes in body composition with advancing age need to be confirmed with longitudinal studies.
251
Mott et al.37 found a highly significant curvilinear relation between age and FM, which indicated a peak amount of FM in late middle age and lower amounts in FM at younger and older ages. Maximum FM was found between 53 and 61 y and maximum %FM from 55 to 71 y. Comparison of %FM determined by BIA with %FM determined by calculations using BMI developed by Deurenberg et al.38 in our study shows progressively larger differences in %FM with advancing age: ⫹0.5% between ages 15 and 24 y and ⫹6.5% between ages 55 and 65 y in men and ⫺0.3% between ages 15 and 24 y and ⫹7.7% between ages 55 and 65 y in women for BMIdetermined %FM. The formula by Deurenberg et al. estimates the increase in %FM to be 2.3% per decade of age. Further confirmation of an effect of age on FM, in addition to an effect of weight gain with age, is necessary. The BIA formulas used in our study do not take age into consideration. Validation of BIA formulas that include an age factor may be necessary for studies that span across 30 – 40 y of age. A recent round table discussion39 suggested that the “best” FM percentages in terms of lowest morbidity and mortality averaged 12–20% in men and 20 –30% in women. Twenty-five percent of all men and 16% of all women in this study were above these recommendations (Table V). Reference Data for Body Composition Parameters The data presented in these tables may prove useful to nutritionists and clinicians who evaluate the health status of healthy and ill subjects and need normal data for comparison. The World Health Organization Expert Committee on Anthropometric Reference Data recommends the collection of data describing local levels and patterns but does not recommend the use of universal reference standards because of worldwide ethnic variations in height, weight, and BMI (and FFM and FM), much of which reflects differences in lifestyle and environment and genetic differences.40 Study Limitations This study included healthy white subjects with different physical activity levels living in Switzerland. Underrepresentation of obese subjects is a possibility because some obese subjects would have been excluded during the recruitment process because of chronic diseases such as diabetes and cardiovascular diseases and are less inclined to be recruited spontaneously in the condition of an open exhibition stand. The BMI of our study was slightly below that of a randomly sampled Swiss population (age 40 –59 y; Euralim Study).23 However, our population included younger subjects (60% of subjects ⬍40 y) and thus proportionally fewer obese subjects. Nineteen percent of subjects had a BMI ⬎25 kg/m2 versus 29.2% reported in the Fourth Nutrition Report in Switzerland (n ⫽ 510). 24 However, this difference may be due to an underrepresentation of older persons (8.5% versus 29% ⬎55 y) in our study and exclusion of obese subjects with secondary pathologies. Thus, underrepresentation of overweight and obese subjects does not appear to be a significant problem. Although these results are valid for healthy, normal subjects, BIA formulas would have to be validated in subjects who deviate from the norm, i.e., body builders, patients with edema, and subjects with significantly different activity levels (very sedentary or extremely physically active). Our study used single-frequency BIA methodology. Bioimpedance spectroscopy would have the advantage of distinguishing the contribution of total body water from extracellular fluid.
CONCLUSION The study presents percentiles for FFM, FM, and %FM of healthy white subjects. Mean FM and %FM increased progressively in
252
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
men and after age 45 y in women. The results suggest that any weight gain noted with advancing age is due to a gain in FM. The percentile data presented serve as references to evaluate the normality of body composition of healthy men and women at a given age. In addition, these reference data may be useful to determine the extent of subnormal body composition in subject groups with various pathologies such as diabetes and hypertension but without fluid and electrolyte abnormalities. These data fill a gap in the literature concerning body composition by providing information on FFM, FM, and %FM and percentile distributions of these parameters in large subject groups. Future studies should be directed toward the quantitative evaluation of FFM and FM as risk factors for chronic diseases such as cardiovascular disease, diabetes, hypertension, and cancer.
ACKNOWLEDGMENTS The authors are indebted to the dietitians at the Geneva University Hospital for data collection and to Khursheed N. Jeejeebhoy for criticism and suggestions for the manuscript.
REFERENCES 1. Roubenoff R, Dallal GE, Wilson PWF. Predicting body fatness: the body mass index vs estimation by bioelectrical impedance. Am J Public Health 1995;85:726 2. Curtin F, Morabia A, Pichard C, Slosman D. Body mass index compared to dual-energy X-ray absorptiometry: evidence for a spectrum bias. J Clin Epidemiol 1997;50:837 3. Heitmann BL. Impedance: a valid method in assessment of body composition? Eur J Clin Nutr 1994;48:228 4. NIH Technology Assessment Conference Statement. Bioelectrical impedance analysis in body composition measurement: National Institutes of Health Technology Assessment Conference Statement. Am J Clin Nutr 1996;64:524S 5. Lukaski HC. Validation of tetrapolar bioelectrical impedance measurements to assess human body composition. J Appl Physiol 1986;60:1327 6. Deurenberg P, Weststrate JA, van der Kooy K. Body composition changes assessed by biolectrical impedance measurements. Am J Clin Nutr 1989;49:401 7. Gray DS. Changes in bioelectrical impedance during fasting. Am J Clin Nutr 1988;48:1184 8. Kushner RF, Schoeller DA. Estimation of total body water by bioelectrical impedance analysis. Am J Clin Nutr 1986;44:417 9. Lukaski HC. Assessment of fat-free mass using bioelectrical impedance measurements of the human body. Am J Clin Nutr 1985;41:810 10. Jackson AS, Pollock ML, Graves JE, Mahar MT. Reliability and validity of bioelectrical impedance in determining body composition. J Appl Physiol 1988; 64:529 11. Segal KR. Use of bioelectrical impedance analysis measurements as an evaluation for participating in sports. Am J Clin Nutr 1996;64:469S 12. Kushner RF, Kunigk A, Alspaugh M, et al. Validation of bioelectrical-impedance analysis as a measurement of change in body composition in obesity. Am J Clin Nutr 1990;52:219 13. Stolarczyk LM, Heyward VH, Van Loan MD, et al. The fatness-specific bioelectrical impedance analysis equations of Segal et al.: are they generalizable and practical? Am J Clin Nutr 1997;66:8 14. Segal KR, Van Loan M, Fitzgerald PI, Hodgdon JA, Van Itallie TB. Lean body mass estimation by bioelectrical impedance analysis: a four-site cross over validation. Am J Clin Nutr 1988;47:7 15. Micozzi MS, Albanes D, Jones DY, Chumlea WC. Correlations of body mass indices with weight, stature, and body composition in men and women in NHANES I and II. Am J Clin Nutr 1986;44:725
16. Frisancho R. New standards of weight and body composition by frame size and height for assessment of nutritional status of adults and the elderly. Am J Clin Nutr 1984;40:808 17. Macdonald SM, Reeder BA, Chen Y, Despres JP. Obesity in Canada: a descriptive analysis. Canadian Heart Health Surveys Research Group. Can Med Assoc J 1997;157:S3 18. Micozzi MS, Harris TM. Age variations in the relation of body mass indices to estimates of body fat and muscle mass. Am J Phys Anthropol 1990;81:375 19. Barlett HL, Puhl SM, Hodgson JL, Buskirk ER. Fat-free mass in relation to stature: ratios of fat-free mass to height in children, adults and elderly subjects. Am J Clin Nutr 1991;53:1112 20. Fentem PH, Mockett SJ. Physical activity and body composition: what do the national surveys reveal? Int J Obesity Rel Metab Disord 1998;22:S8 21. Pichard C, Kyle U, Gremion G, Gerbase M, Slosman D. Body composition by X-ray absorptiometry and bioelectrical impedance in elite female runners. Med Sci Sports Exerc 1997;29:1527 22. Pichard C, Kyle UG. Body composition measurements during wasting diseases. Curr Opin Clin Nutr 1998;1:357 23. Morabia A, Bernstein M, He´ritier S, Ylli A. Community-based surveillance of cardiovascular risk factors in Geneva: methods, resulting distributions, and comparisons with other populations. Prevent Med 1997;26:311 24. Paccaud F, Wietlisbach V, Rickenbach M. Evolution des maladies cardiovasculaires et des caracte´ristiques de l’alimentation: re´sultats de l’e´tude MONICA. In: Office Fe´de´ral de la Sante´ Publique, ed. Quatrie`me Rapport sur la Nutrition en Suisse. Bern: Office Fe´de´ral de la Sante´ Publique, 1998:198 25. Rissanen A, Helio¨vaara M, Aromaa A. Overweight and anthropometric changes in adulthood: a prospective study of 17000 Finns. Int J Obesity 1987;12:391 26. Rolland-Cachera MF, Cole TJ, Sempe´ M, et al. Body mass index variations: centiles from birth to 87 years. Eur J Clin Nutr 1991;45:13 27. Schutz Y, Je´quier E, Office Fe´de´ral de la L’obe´site´. In: Sante´ Publique, ed. Troisie`me Rapport sur la Nutrition en Suisse. Bern: Office Fe´de´ral de la Sante´ Publique, 1991:384 28. Must A, Dallal GE, Dietz WH. Reference data for obesity: 85th and 95th percentiles of body mass index (wt/ht2) and triceps skinfold thickness. Am J Clin Nutr 1991;53:839 29. Flegal KM, Carroll MD, Kuczarski RJ, Johnson CL. Overweight and obesity in the United States: prevalence and trends, 1960 –1994. Int J Obesity 1998;22:39 30. Forbes GB, Reina JC. Adult lean body mass declines with age: some longitudinal observations. Metabolism 1970;19:653 31. Chumlea WC, Guo SS, Zeller CM, Reo NV, Siervogel RM. Total body water data for white adults 18 to 64 years of age: the Fels Longitudinal Study. Kidney Int 1999;56:244 32. Bishop C, Phyllis E, Ritchey S. Norms for nutritionnal assessment of American adults by upper arm anthropometry. Am J Clin Nutr 1981;34:2530 33. Biasioli S, Foroni R, Petrosino L, et al. Effect of aging on the body composition of dialyzed subjects. Comparison with normal subjects. ASAIO J 1993;39:M596 34. Bartlett HL, Puhl SM, Hodgson JL, Burskirk ER. Fat-free mass in relation to stature: ratios of fat-free mass to height in children, adults, and elderly subjects. Am J Clin Nutr 1991;53:1112 35. Kehayias JJ, Fiatarone MA, Zhuang H, Roubenoff R. Total body potassium and body fat: relevance to aging. Am J Clin Nutr 1997;66:904 36. Guo SS, Zeller C, Chumlea WC, Siervogel RM. Aging, body composition, and lifestyle: the Fels Longitudinal Study. Am J Clin Nutr 1999;70:405 37. Mott JW, Wang J, Thornton JC, et al. Relation between body fat and age in 4 ethnic groups. Am J Clin Nutr 1999:1007 38. Deurenberg P, Weststrate JA, Seidell JC. Body mass index as a measure of body fatness: age- and sex-specific prediction formula. Br J Nutr 1991;65:105 39. Abernathy RP, Black DR. Healthy body weights: an alternative perspective. Am J Clin Nutr 1996;63:448S 40. de Onis M, Habicht JP. Anthropometric reference data for international use: recommendations from a World Health Organization Expert Committee. Am J Clin Nutr 1996;64:650
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
253
ANNEX 1. PHYSICAL CHARACTERISTICS (MEAN ⫾ SD) OF VALIDATION GROUPS Non-obese n
Men (80)
Women (69)
Overweight and obese Men (58)
Women (20)
Age (y) 41.2 ⫾ 13.5 43.2 ⫾ 11.7 48.2 ⫾ 10.7 53.3 ⫾ 5.2 Height (cm) 174.7 ⫾ 6.6 160.6 ⫾ 5.7 174.4 ⫾ 6.9 158.8 ⫾ 6.4 Weight (kg) 67.5 ⫾ 9.4 55.0 ⫾ 10.9 86.8 ⫾ 8.8 81.5 ⫾ 8.1 21.2 ⫾ 3.5 28.0 ⫾ 4.1 32.3 ⫾ 1.6 BMI (kg/cm2) 22.1 ⫾ 2.9 IBW (%) 96.4 ⫾ 13.0 94.4 ⫾ 16.1 124.1 ⫾ 10.3 137.9 ⫾ 4.8 FFMDXA (kg) 54.4 ⫾ 6.8 39.3 ⫾ 5.9 63.3 ⫾ 7.0 46.5 ⫾ 4.0 FFMBIA (kg) 54.5 ⫾ 7.5 39.7 ⫾ 6.0 62.9 ⫾ 6.8 45.6 ⫾ 3.9 BMI, body mass index; FFMBIA, fat-free mass measured by bioelectrical impedance analysis; FFMDXA, fat-free mass measured by dual-energy x-ray absorptiometry (DXA); IBW, ideal body weight.
ANNEX 3. Correlations (top) and differences (bottom) of FFM in nonobese women estimated by BIA and dual-energy x-ray absorptiometry (Hologic QDR-4500). The difference (calculated as FFMDXA ⫺ [FFMDXA ⫹ FFMBIA/2 per Bland-Altman]) is plotted against the mean of the measurements of FFM by DXA and BIA (FFMDXA ⫹ FFMBIA/2). Dotted line in top graph is the line of identity. BIA, bioelectrical impedance analysis; DXA, dual-energy x-ray absorptiometry; FFM, fat-free mass; SEE, standard error of the estimate.
ANNEX 2. Correlations (top) and differences (bottom) of FFM in nonobese men estimated by BIA and dual-energy x-ray absorptiometry (Hologic QDR-4500). The difference (calculated as FFMDXA ⫺ [FFMDXA ⫹ FFMBIA/2 per Bland-Altman]) is plotted against the mean of the measurements of FFM by DXA and BIA (FFMDXA ⫹ FFMBIA/2). Dotted line in top graph is the line of identity. BIA, bioelectrical impedance analysis; DXA, dual-energy x-ray absorptiometry; FFM, fat-free mass; SEE, standard error of the estimate.
254
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
ANNEX 4. Correlations (top) and differences (bottom) of FFM in obese men estimated by BIA and dual-energy x-ray absorptiometry (Hologic QDR-4500). The difference (calculated as FFMDXA ⫺ [FFMDXA ⫹ FFMBIA/2 per Bland-Altman]) is plotted against the mean of the measurements of FFM by DXA and BIA (FFMDXA ⫹ FFMBIA/2). Dotted line in top graph is the line of identity. BIA, bioelectrical impedance analysis; DXA, dual-energy x-ray absorptiometry; FFM, fat-free mass; SEE, standard error of the estimate.
ANNEX 5. Correlations (top) and differences (bottom) of FFM in obese women estimated by BIA and dual-energy x-ray absorptiometry (Hologic QDR-4500). The difference (calculated as FFMDXA ⫺ [FFMDXA ⫹ FFMBIA/2 per Bland-Altman]) is plotted against the mean of the measurements of FFM by DXA and BIA (FFMDXA ⫹ FFMBIA/2). Dotted line in top graph is the line of identity. BIA, bioelectrical impedance analysis; DXA, dual-energy x-ray absorptiometry; FFM, fat-free mass; SEE, standard error of the estimate.
Reference Values of Fat-Free and Fat Masses by Bioelectrical Impedance Analysis in 3393 Healthy Subjects Claude Pichard, MD, PhD, Ursula G. Kyle, MS, RD, David Bracco, MD, Daniel O. Slosman, MD, Alfredo Morabia, MD, PhD, and Yves Schutz, PhD From the Divisions of Clinical Nutrition, Nuclear Medicine, and Epidemiology, Geneva University Hospital, Geneva; and the Institute of Physiology, Lausanne, Switzerland Determination of fat-free mass (FFM) and fat mass (FM) is of considerable interest in the evaluation of nutritional status. In recent years, bioelectrical impedance analysis (BIA) has emerged as a simple, reproducible method used for the evaluation of FFM and FM, but the lack of reference values reduces its utility to evaluate nutritional status. The aim of this study was to determine reference values for FFM, FM, and %FM by BIA in a white population of healthy subjects, to observe the changes in these values with age, and to develop percentile distributions for these parameters. Whole-body resistance of 1838 healthy white men and 1555 women, aged 15– 64 y, was determined by using four skin electrodes on the right hand and foot. FFM and FM were calculated according to formulas validated for the subject groups and analyzed for age decades. This is the first study to present BIA-determined age- and sex-specific percentiles for FFM, FM, and %FM for healthy subjects, aged 15– 64 y. Mean FM and %FM increased progressively in men and after age 45 y in women. The results suggest that any weight gain noted with age is due to a gain in FM. In conclusion, the data presented as percentiles can serve as reference to evaluate the normality of body composition of healthy and ill subject groups at a given age. Nutrition 2000;16:245–254. ©Elsevier Science Inc. 2000 Key words: bioelectrical impedance analysis, fat-free mass, body fat, fat-free mass measured by bioelectrical impedance analysis, body composition, reference standard, gender
INTRODUCTION Determination of fat-free mass (FFM) and fat mass (FM) is of considerable interest in the evaluation of nutritional status. Both over- and undernutrition contribute to increased mortality and morbidity. The most common estimate of excess or deficit in body weight is the body mass index (BMI). However, BMI has been shown to be an imprecise measurement of fatness.1,2 Simple measurements for evaluating body composition such as skinfold measurements are easy to perform but lack accuracy and reproducibility.1 Other methods require subject cooperation (underwater weighing) or sophisticated equipment and skilled technicians (tracer dilution, dual-energy x-ray absorptiometry (DXA), and neutron activation analysis). To date, the only measures used to grade body composition in very large studies (such as the Study by Metropolitan Life Insurance) have been relative weight, weight for height, or BMI. In recent years, bioelectrical impedance analysis (BIA) has emerged as a simple, reproducible method that can be used for the evaluation of FFM and FM. Because of its low cost and ease and rapidity of use, one of the greatest potential uses of BIA is in studies of large populations. BIA represents a substantial improvement in the epidemiologic method of body composition compared with BMI1; it seems to be valid for subjects without large disturbances in fluid distribution and provides a more reliable measure-
Correspondence to: Claude Pichard, MD, PhD, Clinical Nutrition and Diet-therapy, Geneva University Hospital, 1211 Geneva, Switzerland. Email: [email protected] Date accepted: Jan. 19, 2000. Nutrition 16:245–254, 2000 ©Elsevier Science Inc., 2000. Printed in the United States. All rights reserved.
ment of body composition with respect to FFM and FM than do simpler methods such as skinfolds and height and weight.3,4 The aims of this study were to determine reference values for FFM, FM, and %FM by BIA in a large white (Western European) population of healthy subjects in Switzerland, to observe differences in these values in age groups of 10 y between ages 15 and 64 y, and to develop percentile distributions for these parameters. This is the first study that examines FFM and FM by BIA in an adult population across the age spectrum and can serve as a baseline to monitor future trends in body composition and changes in FFM, FM, and %FM. Moreover, age- and sex-specific percentile distributions, not available until the present time, will permit more meaningful interpretations of FFM and FM. The data presented as percentiles can serve as references to evaluate the normality of body composition of healthy or ill groups at a given age.
SUBJECTS AND METHODS Subjects To determine reference values for a healthy subject population, 3393 healthy adults (1838 men and 1555 women), aged 15– 64 y, were recruited by offering free BIA on an exhibition stand at trade fairs and fun runs and among public administration staff. This method allowed us to include a wide representation of people with different levels of physical activity. The anthropometric data and number of healthy subjects per age group is shown in Table I. All subjects were ambulatory, with no known pathologies or physical handicaps. Subjects were questioned on use of medications and reasons for visits to a physician in the last 6 mo to eliminate subjects with acute or chronic diseases. Due to the heterogeneity of 0899-9007/00/$20.00 PII S0899-9007(00)00256-2
246
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS TABLE I. ANTHROPOMETRIC AND BIOIMPEDANCE CHARACTERISTICS (MEAN ⫾ SD) OF A HEALTHY WHITE POPULATION
Age (y) Men n Height (cm) Weight (kg) BMI (kg/m2) IBW (%) Resistance (⍀) Reactance (⍀) Women n Height (cm) Weight (kg) BMI (kg/m2) IBW (%) Resistance (⍀) Reactance (⍀)
15–64*
15–24
25–34
35–44
45–54
55–64
1838 176.3 ⫾ 6.7 (155–199) 73.2 ⫾ 9.2 (48.1–117.6) 23.5 ⫾ 2.7 (17.1–40.1) 108.4 ⫾ 13.0 (74–172) 475 ⫾ 48 (334–743) 60.7 ⫾ 9.9 (28.8–113)
368 177.5 ⫾ 6.5 69.7 ⫾ 8.2 22.1 ⫾ 2.1 101.9 ⫾ 10.3 483 ⫾ 53 63.3 ⫾ 9.5
503 177.6 ⫾ 6.6 73.3 ⫾ 88 23.2 ⫾ 2.3 106.2 ⫾ 12.1 475 ⫾ 47 62.8 ⫾ 10.5
489 176.6 ⫾ 6.5 74.7 ⫾ 9.5 24.0 ⫾ 2.8 110.1 ⫾ 13.1 472 ⫾ 50.2 60.0 ⫾ 10.1
306 174.1 ⫾ 6.3 73.9 ⫾ 9.2 24.4 ⫾ 2.6 112.7 ⫾ 11.5 473 ⫾ 42 58.9 ⫾ 8.5
172 173.4 ⫾ 6.5 75.3 ⫾ 9.5 25.0 ⫾ 3.0 116.6 ⫾ 14.0 472 ⫾ 46 55.3 ⫾ 11.1
1555 164.5 ⫾ 6.3 (147–186) 58.9 ⫾ 8.2 (39.1–99.8) 21.8 ⫾ 2.9 (16.8–37.7) 105.1 ⫾ 13.7 (73–165) 580 ⫾ 61 (414–837) 67.4 ⫾ 13.8 (32–132)
407 166.5 ⫾ 6.2 57.7 ⫾ 7.7 20.8 ⫾ 2.4 101.0 ⫾ 12.1 601 ⫾ 61 72.0 ⫾ 14.2
425 165.4 ⫾ 6.1 58.3 ⫾ 7.8 21.3 ⫾ 2.5 101.8 ⫾ 13.1 582 ⫾ 61 66.5 ⫾ 13.8
344 163.6 ⫾ 5.9 58.7 ⫾ 8.1 21.9 ⫾ 2.7 105.8 ⫾ 13.0 570 ⫾ 55 67.7 ⫾ 15.1
247 163.2 ⫾ 5.9 60.0 ⫾ 8.4 22.5 ⫾ 2.8 109.3 ⫾ 13.0 562 ⫾ 58 64.4 ⫾ 13.6
132 160.5 ⫾ 5.9 63.5 ⫾ 9.4 24.7 ⫾ 3.6 114.2 ⫾ 15.1 564 ⫾ 65 61.7 ⫾ 13.7
* Ranges are given in parentheses. BMI, body mass index; IBW, ideal body weight (Metropolitan Life Insurance, 1983).
the population (about one-third of population is of non-Swiss nationality), the sample is more representative of a European than strictly Swiss population. Subjects older than 65 y were excluded because of insufficient validation of BIA formulas in elderly subjects. The study protocol complied with the requirements of the Geneva University Hospital Ethics Rules. Anthropometric Measurements and Bioelectrical Impedance Analysis Body height was measured to the nearest 0.5 cm and body weight to the nearest 0.1 kg on a balance beam scale (Seca Corp. Mechanical Scale, Berlin, Germany). Resistance and reactance were measured by a BIA generator and used to mathematically derive FFM and FM, as previously described,5–7 by using the formula V ⫽ ⫻ ht2/R, in which conductive volume (V) is assumed to represent FFM, is the specific resistivity of the conductor, height (ht) is taken as the length of the conductor, and body resistance (R) is measured with four surface electrodes placed on the right wrist and ankle. Briefly, an electrical current of 50 kHz and 0.8 mA was produced by a generator (Bio-Z2, Spengler, Paris, France) and applied to the skin with adhesive electrodes (3M Red Dot, 3M Health Care, Borken, Germany) with the subject in the decubitus dorsalis.8 The skin was cleaned with 70% alcohol. The technical range of the Bio-Z2 generator is: impedance, 100 to 900 ⍀; phase angle, 1 to 18°; and an accuracy of ⫾3 ⍀ and ⫾0.2°, respectively. Short- and long-term reproducibility of resistance measurements indicate coefficients of variance of 1.8 –2.9%.9,10 Several BIA instruments were used in this study to permit inclusion of a large number of subjects at one time. The Bio-Z2 generators were cross-validated at 50 kHz against the RJL-109 and 101 analyzers (RJL Systems Inc., Clinton Township, MI, USA) and against the Xitron analyzer (Xitron Technologies, Inc., San Diego, CA, USA). Limit of tolerance between instruments was ⫾5 ⍀ at 50 kHz using a calibration jig and in vivo measurements. All data collectors had standard training and were provided verbal and written information to standardize the procedure and minimize errors due to multiple data collectors. All measurements were made 2–3 h after meal intake, and measurements of participants in fun runs were made before the run
to avoid changes in hydration, skin temperature, electrolyte concentration, and glycogen stores.11 Validation of BIA Formulas Body composition was calculated with formulas shown in Table II. Validation of FFM against DXA (Hologic QDR-4500, Whole Body software version, Hologic Inc., Waltham, MA, USA) in subgroup of 80 non-obese men and 69 non-obese women resulted in a correlation of r ⫽ 0.97 and 0.96, standard error of the estimate (SEE) of 2.0 and 1.6 kg, coefficient of variation (CV) of 3.7% and 4.1%, and mean difference of ⫺0.1 ⫾ 2.0 kg and ⫺0.4 ⫾ 1.6 kg in men and women, respectively (unpublished data), confirming the appropriateness of the BIA formula in the
TABLE II. FORMULAS TO CALCULATE FAT-FREE MASS USING BIOELECTRICAL IMPEDANCE ANALYSIS* Body density† %Fat mass FFM Overweight FFM‡ Women %Fat mass†
FFM Overweight FFM‡
1.1554 ⫺ (0.0841 ⫻ weight ⫻ resistance/[height2]) (4.95/body density ⫺ 4.5) ⫻ 100 Weight ⫺ fat mass BMI ⬎26 kg/m2 14.5244 ⫹ 0.000886 ⫻ (height2) ⫺ 0.02299 ⫻ R ⫹ 0.42688 ⫻ weight ⫺ 0.07002 ⫻ age (1 ⫺ [0.3981 ⫻ height2/resistance ⫹ 0.3066 ⫻ weight ⫹ 0.095299 ⫻ {height ⫺ 100} ⫹ 0.7414]/weight) ⫻ 100 Weight ⫺ fat mass BMI ⬎ 30 kg/m2 9.3794 ⫹ 0.000912 ⫻ (height2) ⫺ 0.01466 ⫻ R ⫹ 0.2999 ⫻ weight ⫺ 0.07012 ⫻ age
* Different formulas were used for overweight subjects. † Manufacturer’s formula (RJL Systems Inc.). ‡ From Segal et al.14 BMI, body mass index; FFM, fat-free mass.
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
247
TABLE III. PERCENTILES FOR FAT-FREE MASS (kg) BY BIOELECTRICAL IMPEDANCE ANALYSIS FOR WHITE ADULTS Percentile Age (y) group Men Total 15–24 25–34 35–44 45–54 55–64 Women Total 15–24 25–34 35–44 45–54 55–64
n
Mean ⫾ SD
5th
10th
25th
50th
75th
90th
95th
1838 368 503 489 306 172
60.3 ⫾ 6.2 59.5 ⫾ 6.7 61.1 ⫾ 6.3* 61.0 ⫾ 6.2 59.3 ⫾ 5.4* 59.1 ⫾ 5.6
50.3 48.2 50.7 51.6 50.2 51.0
52.5 50.9 52.9 53.3 52.5 52.8
55.8 55.0 56.6 56.5 55.8 55.0
60.4 59.7 61.6 60.9 59.0 59.0
64.4 63.9 65.0 65.3 62.8 62.6
68.1 68.0 68.7 69.3 66.3 66.6
70.7 70.4 71.2 71.6 69.4 69.3
1555 407 425 344 247 132
43.7 ⫾ 4.5 43.3 ⫾ 4.7 43.7 ⫾ 4.4 43.6 ⫾ 4.5 44.1 ⫾ 4.4 44.0 ⫾ 4.5
36.6 35.8 36.6 37.0 36.6 37.0
37.8 37.3 38.1 37.8 37.8 38.3
40.5 40.2 40.4 40.5 41.6 40.3
43.6 42.9 43.7 43.7 44.3 43.8
46.6 46.2 46.7 46.6 46.6 47.1
49.5 49.6 49.0 49.8 49.5 49.6
51.4 51.4 51.4 51.1 51.6 51.5
* P ⬍ 0.001 versus preceeding age group, analysis of variance.
subjects of this study. Results of validation are shown in Annex 1 (anthropometric data). Simple regressions and Bland-Altman analysis of non-obese subjects are shown in Annexes 2 (men) and 3 (women). Several studies12–14 have shown that BIA formulas developed for normal-weight subjects are not adapted to overweight subjects. Therefore, we used formulas by Segal et al.14 for overweight subjects (Table II) to calculate FFM in men with BMI greater than 26 kg/m2 and in women with BMI greater than 30 kg/m2. These cutoff values were obtained from our validation of BIA against DXA (unpublished data). The calculations by Segal et al. were used in 291 (15.8%) men and 53 (3.4%) women who fell in the overweight/obese category. Validation of FFM against DXA in 58 obese men and 20 obese women resulted in a correlation of r ⫽ 0.95 and 0.82, SEE of 2.2 and 3.0 kg, CV of 3.6% and 5.1%, and mean difference of 0.4 ⫾ 2.3 kg and 0.9 ⫾ 2.4 kg in men and women, respectively (unpublished data), confirming the appropriateness of the BIA formula in the obese subjects. Simple regressions and Bland-Altman analysis of validation in obese subjects are shown in Annexes 4 (men) and 5 (women).
Statistics The statistical analysis program StatView, version 4.1 (Abacus Concepts, Berkeley, CA, USA) was used for statistical analysis. The results are expressed as mean ⫾ standard deviation. Age- and sex-specific percentile distributions were calculated for each of the following parameters: FFM, FM, and %FM. To compare our results with those of other studies, the data were stratified by steps of 10 y as reported for BMI and anthropometric data in the NHANES study15,16 and a Canadian study.17 The ith percentile (Pi) was the value at or below which there was i% of the sample. For example, the 50th percentile (P50) was the value at or below which there were 50% of the observations for a given variable. Given a total of n ordered values for each parameter ( x 1 , x 2 , x 3 , . . . , x n ), Pi in any of the calculated distribution was computed as follows with the Statview 4.1 statistical program: Pi ⫽ (1 ⫺ A) ( x b ) ⫹ ( A) ( x b ⫹ 1 ). The differences between age groups were analyzed by analysis of variance with Fisher’s protected least-significant-difference comparison.
RESULTS Table I shows the anthropometric and bioimpedance characteristics of the men and women with no known pathologies. Age- and sex-specific percentile distributions for FFM, FM, and %FM of these adults are presented in Tables III–V and in Figures 1–2. Body Mass Index The mean BMI increased from 23.2 kg/m2 in men and 21.3 kg/m2 in women aged 25–34 y to 25.0 kg/m2 in men and 24.7 kg/m2 in women aged 55– 64 y (Table I). Fat-Free Mass The mean FFM (Table III) was 60.3 ⫾ 6.2 kg for white men between 15 and 64 y. Of these men, 90% had an FFM between 52.5 and 68.1 kg. The mean FFM was greatest for men aged 25– 44 y. Mean FFM was significantly higher in men aged 25–34 y than in men aged 15–24 y and significantly lower in men aged 45–54 y than in men aged 35– 44 y. The percentiles developed for men between 25 and 44 y generally exceeded comparable percentiles calculated for men of other age groups. A diagram of these trends is presented in Figure 1. For women, the mean FFM was 43.7 ⫾ 4.5 kg. Ninety percent of women had an FFM between 37.8 and 49.5 kg. The FFM increased slightly but non-significantly from 43.3 ⫾ 4.7 kg for women aged 15–24 y to 44.1 ⫾ 4.4 kg for women aged 45–54 y. These age-related trends were also apparent in the percentile distributions. Figure 1 displays these trends in graphic form. Fat Mass and Percentage of Fat Mass The mean FM (Table IV) for white men between the ages of 15 and 64 y was 12.9 ⫾ 5.7 kg. Of these men, 90% had an FM between 7.4 and 22.1 kg. FM of men showed differences related to aging. The FM increase was significant for all age groups. The mean FM was 11.1% greater in those 35– 44 y old and 32.8% greater in those 55– 64 y old than in those 25–34 y old. The mean %FM (Table V) for white men was 17.3 ⫾ 5.8%. Of these men, 90% had a %FM between 11.1 and 27.1%. The parallel differences
248
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS TABLE IV. PERCENTILES FOR FAT MASS (kg) BY BIOELECTRICAL IMPEDANCE ANALYSIS FOR WHITE ADULTS Percentile
Age (y) group Men Total 15–24 25–34 35–44 45–54 55–64 Women Total 15–24 25–34 35–44 45–54 55–64
n
Mean ⫾ SD
5th
10th
25th
50th
75th
90th
95th
1838 368 503 489 306 172
12.9 ⫾ 5.7 10.2 ⫾ 3.7 12.2 ⫾ 4.7* 13.6 ⫾ 5.8* 14.5 ⫾ 6.0† 16.2 ⫾ 7.2*
6.6 6.1 6.6 6.8 7.4 7.0
7.4 6.8 7.4 7.7 8.3 8.1
9.0 7.9 8.9 9.6 9.9 10.6
11.3 9.4 11.2 12.1 13.1 14.3
15.1 11.4 14.3 16.5 18.3 22.7
22.1 14.5 19.3 22.2 23.5 26.6
24.5 17.1 22.3 24.8 26.2 27.5
1555 407 425 344 247 132
15.2 ⫾ 5.0 14.3 ⫾ 4.0 14.6 ⫾ 4.6 15.1 ⫾ 4.9 15.8 ⫾ 5.3 19.5 ⫾ 6.5*
9.0 8.8 8.9 9.0 8.8 11.2
9.9 9.9 9.6 9.9 9.7 12.8
11.9 11.7 11.6 12.0 12.3 15.1
14.4 13.8 14.0 14.2 15.6 19.1
17.4 16.4 16.7 17.1 19.1 21.9
21.0 19.2 19.4 21.2 21.7 26.0
23.5 21.0 21.7 22.4 23.2 31.7
* P ⬍ 0.001 versus preceeding age group, analysis of variance. † P ⬍ 0.05 versus preceeding age group, analysis of variance.
between age groups were not consistently present in the percentile distribution. Greater dispersion was noted in the upper percentiles and older subjects. The mean FM for women (Table IV) between the ages of 15 and 64 y was 15.2 ⫾ 5.0 kg. Of these women, 90% had an FM between 9.9 and 21.0 kg. Marked differences were observed between the FM means determined for the group older than 55 y. The largest mean, observed in those aged 55– 64 y, exceeded the smallest mean, noted in those aged 15–24 y, by 36.4%. The mean %FM for white women was 25.3% (Table V). Of these women, 90% had an FM between 19.2 and 31.6%. A significant increase was only noted in %FM in women older than 55 y. Percentile distributions were parallel in women aged 15–54 y, but showed marked differences between age groups in those older than 55 y,
with increases noted in all percentile distributions. Figure 2 shows trends in FM, and Figure 3 shows trends in %FM for men and women.
Gender Differences in Fat-Free Mass and Fat Mass As expected, gender differences were noted in FFM. The mean FFM was 38% greater in men than in women and was slightly higher in 25–34-y-old subjects than in those older than 45 y. The age distribution showed that the mean FFM remained in a narrow range of 59.1– 61.0 kg for men and 43.3– 44.1 kg for women between ages 15 and 64 y. The higher FFM in men and relatively stable FFM throughout the adult age range is confirmed by a mean
TABLE V. PERCENTILES FOR FAT MASS (%) BY BIOELECTRICAL IMPEDANCE ANALYSIS FOR WHITE ADULTS Percentile Age (y) group Men Total 15–24 25–34 35–44 45–54 55–64 Women Total 15–24 25–34 35–44 45–54 55–64
n
Mean ⫾ SD
5th
10th
25th
50th
75th
90th
95th
1838 368 503 489 306 172
17.3 ⫾ 5.8 14.5 ⫾ 4.3 16.3 ⫾ 4.9* 17.8 ⫾ 5.8* 19.2 ⫾ 6.0* 20.9 ⫾ 7.2*
10.1 9.4 10.1 10.3 11.3 10.3
11.1 10.4 10.9 11.7 12.4 11.7
13.1 11.9 12.9 13.5 14.7 15.2
15.8 13.7 15.3 16.4 17.7 19.1
20.0 16.1 18.8 21.4 24.5 28.4
27.1 19.3 24.0 26.9 28.5 30.7
28.8 22.5 26.2 28.7 29.7 32.0
1555 407 425 344 247 132
25.4 ⫾ 5.1 24.5 ⫾ 4.2 24.6 ⫾ 4.7 25.2 ⫾ 5.1 25.9 ⫾ 5.5 30.1 ⫾ 5.8*
17.6 18.1 17.4 17.7 16.9 22.3
19.2 19.6 18.8 19.1 18.2 24.3
22.0 21.7 21.5 21.9 22.4 26.1
25.1 24.4 24.4 24.8 26.4 30.0
28.3 27.1 27.2 28.2 29.5 32.8
31.6 29.8 30.2 31.6 31.9 36.0
33.4 31.3 31.9 32.9 33.6 42.6
* P ⬍ 0.001 versus the preceeding age group, analysis of variance.
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
249
FIG. 1. Apparent changes in fat-free mass of white adults with advancing age.
FFM index (kg FFM/ht2; data not shown) of 19.3 kg/cm2 in 25–34-y-old men and 19.6 kg/cm2 in 35– 64-y-old men and of 16.0, 16.3, 16.5, and 17.0 kg/m2 in 25–34, 35– 44, 45–54, and 55– 64-y-old women, respectively. Gender differences were also noted in FM and %FM. The higher FM in women is confirmed by higher mean FM indexes (kg FM/ht2) of 5.2, 5.3, 5.6, 5.9, and 7.6 kg/m2 in women and 3.2, 3.8, 4.3, 4.8, and 5.4 kg/m2 in men aged 15–24, 25–34, 35– 44, 45–54, and 55– 64 y, respectively. FM indexes increased significantly (P ⬍ 0.01) in all age groups except in women 25–34 y versus those 15–24 y and in women 55– 64 y versus those 45–54 y. A larger distribution in %FM was noted in women. Mean %FM was
FIG. 2. Apparent changes in fat mass of white adults with advancing age.
significantly greater in women than in men. The 95th percentile for %FM was 42.6% for women versus 32.0% for men, suggesting a greater increase with age in %FM in women than in men. Furthermore, larger FM and %FM in older adults showed that any weight gain was explained by FM gain in both sexes.
DISCUSSION Few studies have reported on FFM and %FM in large population samples,1,18,19 and none have reported on sex- and age-specific percentiles for measured FFM, FM, and %FM. Most studies that
250
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
FIG. 3. Apparent changes in percentage of fat mass in white adults with advancing age.
have reported on FM are based on skinfold measurements and other indirect estimations such as waist circumference and waistto-hip ratios.17,20 BIA is being recognized as an easy, portable method that allows estimation of body composition compartments. Despite concerns about the validity of BIA when compared with sophisticated reference techniques, it remains the only technique that is inexpensive and applicable to large numbers of subjects.1,4 As far as we know, ours is the first study to examine FFM and FM by BIA in a large, white, healthy population and evaluate body composition parameters across the age spectrum from 15 to 64 y. The percentiles derived from these results can serve as references for clinical evaluation of body composition in healthy and ill subjects. Validation of BIA Formulas Validity of FFM and FM obtained by BIA depends on the use of an equation appropriate for the subset of the population.1,21,22 Different BIA formulas were used in men and women and in overweight and normal-weight subjects. These formulas were validated (see SUBJECTS AND METHODS and Annexes 1 through 5) against DXA. The validation confirmed the appropriateness of the BIA formula in healthy subjects participating in the present study, including overweight subjects. Body Mass Index The BMI in this study serves to compare the subjects with those in other epidemiologic studies. Because BMI is a rough indicator of fatness, studies including subjects with higher BMI, such as the Framingham studies in the United States,1 would be expected to report higher %FM than that reported in this study. The BMI of our study was slightly lower than the BMI of 25.3 kg/m2 in men and 23.0 kg/m2 in women reported a randomly sampled Swiss population (ages 40 –59 y) evaluated as part of a community-based surveillance program (the Euralim Study, a European public-health study of eating habits and cardiovascular disease and cancer risk factors).23 The difference is due to our population including younger subjects (60% of subjects ⬍40 y) and proportionally fewer obese subjects. Nineteen percent of subjects had a BMI greater than 25 kg/m2 compared with 29.2%
reported in the Fourth Nutrition Report in Switzerland (n ⫽ 510). 24 However, this difference may be due to an underrepresentation of older persons (8.5% versus 29% ⬎55 y) in our study and the exclusion of obese subjects with secondary pathologies. The BMI in this study is comparable to a mean BMI of 24.6 ⫾ 3.6 kg/m2 in men and 25.3 kg/m2 in women in Finland25 and a BMI of 21.5 kg/m2 at age 20 y and 25.4 kg/m2 at age 60 y in men and a BMI of 20.6 kg/m2 at age 20 y and 24.1 kg/m2 at age 60 y in women in France.26 Thus, mean BMI in Swiss subjects is similar to that in other European populations.27 The BMI was lower than BMIs of 28.6 ⫾ 4.4 kg/m2 in men aged 30 –39 y and 29.0 ⫾ 3.5 kg/m2 in men aged 60 – 69 y and BMIs of 27.2 kg/m2 in women aged 30 to 39 y and 28.0 ⫾ 5.7 kg/m2 in women aged 60 – 69 y reported by the Framingham Offspring Study in the United States.1 Important differences in BMI are noted in men younger than 54 y and in women younger than 64 y when compared with the large, nationally representative North American white sample,28 which had BMIs 4 –7 kg/m2 higher at the 90th percentile than the white subjects in this study.27 Recent studies have suggested significant increases in mean BMI in North Americans in the last 20 y.29 Such an increase is not apparent in Switzerland,24 although representative data of the whole population are lacking. Fat-Free Mass No study has reported changes of FFM in large population samples over several decades. Progressive decline in FFM with advancing age reported in a longitudinal study30 could not be confirmed in the present cross-sectional study. Bartlett et al.19 noted a decline beginning at age 60 y in men. Mean FFM decreased slightly but was statistically significant in men aged 45–54 y versus men aged 35– 44 y in our study, but this mean decrease in FFM with age was not significant when corrected for the shorter stature of older men. The earlier slight decrease in FFM in women, from 51–55 y, reported by Bartlett et al.19 was not confirmed in our study. In women, FFM normalized for height shows a slight increase in FFM until age 64 y, probably related to heavier subjects having greater FFM. Bartlett et al.19 determined FFM and FM by underwater weighing in different age groups consisting of approximately 1000 healthy subjects. They found similar FFM (60.0 – 62.2 kg) in men
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS aged 20 – 60 y and higher FFM (45.0 – 47.6 kg versus 43.3– 44.1 kg in the present study) in women who had higher BMIs, thus confirming FFM as being similar in different populations as measured by different body composition methods. The Fels Longitudinal Study31 reported 59.4 kg of FFM (by DXA) in men aged 20 –29 y and 60.2 kg in men older than 60 y and 42.2 kg of FFM in women aged 20 –29 y, 41.1 kg in women aged 50 –59 y, and 38.3 kg in women older than 60 y; women older than 60 y had a lower BMI than did women aged 50 to 59 y, suggesting that FFM remains stable if body mass does not decrease. A drop in FFM usually occurs after age 65 y, an age category not included in this study because of lack of adequate validation of BIA formulas in elderly subjects. The results of this study suggest that the mean FFM remains relatively stable in healthy adults throughout adulthood until approximately 60-y-old. The average body mass was progressively higher in each category of age (about ⫹5 kg in 40 y).27 Because body mass gain is accompanied by an increase in both FM and FFM, this increase will offset the drop in FFM with advancing age expected at a normal body weight. Fat Mass and Percentage of Fat Mass The results of this study showed that body weight and BMI increased until age 64 y and that the increase was predominantly due to an increase in the FM. Numerous studies have reported increases in subscapular and triceps skinfolds, which represent FM, with advancing age.25,32 This study agrees with previous reports and shows a significant increase in FM and %FM in men in all age ranges, which is parallel to increases in weight and a trend of an increase in FM and %FM in women, but which reached statistical significance only in women 55– 64-y-old as opposed to women 45–54-y-old. Because the data are cross-sectional rather than longitudinal, trends in body composition parameters observed with advancing age may have represented differences between successive generations as well as physiologic alterations with aging. Roubenoff et al.1 found higher %FM (29.9 ⫾ 6.1% in men and 37.7 ⫾ 7.8% in women) in subjects with higher BMIs participating in the Framingham studies and noted a mean increase in %FM with advancing age in both sexes, peaking in the 60s and 50s for American women and men, respectively, which is inconsistent with our results of an increase in %FM in men until age 64 y. Biasoli et al.33 reported similar %FMs by BIA in a small number of subjects with higher BMIs. The relative lower FM may be related to using the same formula for overweight subjects; thus, FM may have been underestimated in their more obese subjects. Higher %FM by hydrodensitometry in men and women with higher weights and BMIs than those in our study have been reported by Bartlett et al.34 The difference may also have been due to inclusion in our study of some subjects with relatively high physical activity levels and thus lower FM. Keharias et al.35 also found higher mean %FM (ranging from 14.9 –25.7% in men and from 28.0 –35.4% in women) by neutron activation analysis in subjects with higher BMIs. Similarly, the Fels Longitudinal Study31 reported 19.5%FM in men aged 20 –29 y with a mean BMI of 23.1 kg/m2 and 31.2%FM in men older than 60 y with a BMI of 29.8 kg/m2 and 34%FM in women aged 20 –29 y with a BMI of 23.6 kg/m2 and 39.4%FM in women older than 60 y with a mean BMI of 24.4 kg/m2. Comparisons with other studies in which height, weight, and age are significantly different are difficult to make. Guo et al.36 reported higher FM and %FM by underwater weighing in men (n ⫽ 102) and women (n ⫽ 109) aged 40 – 66 y who had higher BMIs than those reported in our study and found a significant decrease in FFM and increases in total FM, %FM, body weight, and BMI with advancing age in their longitudinal study (mean follow-up of 9 y). Thus, changes in body composition with advancing age need to be confirmed with longitudinal studies.
251
Mott et al.37 found a highly significant curvilinear relation between age and FM, which indicated a peak amount of FM in late middle age and lower amounts in FM at younger and older ages. Maximum FM was found between 53 and 61 y and maximum %FM from 55 to 71 y. Comparison of %FM determined by BIA with %FM determined by calculations using BMI developed by Deurenberg et al.38 in our study shows progressively larger differences in %FM with advancing age: ⫹0.5% between ages 15 and 24 y and ⫹6.5% between ages 55 and 65 y in men and ⫺0.3% between ages 15 and 24 y and ⫹7.7% between ages 55 and 65 y in women for BMIdetermined %FM. The formula by Deurenberg et al. estimates the increase in %FM to be 2.3% per decade of age. Further confirmation of an effect of age on FM, in addition to an effect of weight gain with age, is necessary. The BIA formulas used in our study do not take age into consideration. Validation of BIA formulas that include an age factor may be necessary for studies that span across 30 – 40 y of age. A recent round table discussion39 suggested that the “best” FM percentages in terms of lowest morbidity and mortality averaged 12–20% in men and 20 –30% in women. Twenty-five percent of all men and 16% of all women in this study were above these recommendations (Table V). Reference Data for Body Composition Parameters The data presented in these tables may prove useful to nutritionists and clinicians who evaluate the health status of healthy and ill subjects and need normal data for comparison. The World Health Organization Expert Committee on Anthropometric Reference Data recommends the collection of data describing local levels and patterns but does not recommend the use of universal reference standards because of worldwide ethnic variations in height, weight, and BMI (and FFM and FM), much of which reflects differences in lifestyle and environment and genetic differences.40 Study Limitations This study included healthy white subjects with different physical activity levels living in Switzerland. Underrepresentation of obese subjects is a possibility because some obese subjects would have been excluded during the recruitment process because of chronic diseases such as diabetes and cardiovascular diseases and are less inclined to be recruited spontaneously in the condition of an open exhibition stand. The BMI of our study was slightly below that of a randomly sampled Swiss population (age 40 –59 y; Euralim Study).23 However, our population included younger subjects (60% of subjects ⬍40 y) and thus proportionally fewer obese subjects. Nineteen percent of subjects had a BMI ⬎25 kg/m2 versus 29.2% reported in the Fourth Nutrition Report in Switzerland (n ⫽ 510). 24 However, this difference may be due to an underrepresentation of older persons (8.5% versus 29% ⬎55 y) in our study and exclusion of obese subjects with secondary pathologies. Thus, underrepresentation of overweight and obese subjects does not appear to be a significant problem. Although these results are valid for healthy, normal subjects, BIA formulas would have to be validated in subjects who deviate from the norm, i.e., body builders, patients with edema, and subjects with significantly different activity levels (very sedentary or extremely physically active). Our study used single-frequency BIA methodology. Bioimpedance spectroscopy would have the advantage of distinguishing the contribution of total body water from extracellular fluid.
CONCLUSION The study presents percentiles for FFM, FM, and %FM of healthy white subjects. Mean FM and %FM increased progressively in
252
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
men and after age 45 y in women. The results suggest that any weight gain noted with advancing age is due to a gain in FM. The percentile data presented serve as references to evaluate the normality of body composition of healthy men and women at a given age. In addition, these reference data may be useful to determine the extent of subnormal body composition in subject groups with various pathologies such as diabetes and hypertension but without fluid and electrolyte abnormalities. These data fill a gap in the literature concerning body composition by providing information on FFM, FM, and %FM and percentile distributions of these parameters in large subject groups. Future studies should be directed toward the quantitative evaluation of FFM and FM as risk factors for chronic diseases such as cardiovascular disease, diabetes, hypertension, and cancer.
ACKNOWLEDGMENTS The authors are indebted to the dietitians at the Geneva University Hospital for data collection and to Khursheed N. Jeejeebhoy for criticism and suggestions for the manuscript.
REFERENCES 1. Roubenoff R, Dallal GE, Wilson PWF. Predicting body fatness: the body mass index vs estimation by bioelectrical impedance. Am J Public Health 1995;85:726 2. Curtin F, Morabia A, Pichard C, Slosman D. Body mass index compared to dual-energy X-ray absorptiometry: evidence for a spectrum bias. J Clin Epidemiol 1997;50:837 3. Heitmann BL. Impedance: a valid method in assessment of body composition? Eur J Clin Nutr 1994;48:228 4. NIH Technology Assessment Conference Statement. Bioelectrical impedance analysis in body composition measurement: National Institutes of Health Technology Assessment Conference Statement. Am J Clin Nutr 1996;64:524S 5. Lukaski HC. Validation of tetrapolar bioelectrical impedance measurements to assess human body composition. J Appl Physiol 1986;60:1327 6. Deurenberg P, Weststrate JA, van der Kooy K. Body composition changes assessed by biolectrical impedance measurements. Am J Clin Nutr 1989;49:401 7. Gray DS. Changes in bioelectrical impedance during fasting. Am J Clin Nutr 1988;48:1184 8. Kushner RF, Schoeller DA. Estimation of total body water by bioelectrical impedance analysis. Am J Clin Nutr 1986;44:417 9. Lukaski HC. Assessment of fat-free mass using bioelectrical impedance measurements of the human body. Am J Clin Nutr 1985;41:810 10. Jackson AS, Pollock ML, Graves JE, Mahar MT. Reliability and validity of bioelectrical impedance in determining body composition. J Appl Physiol 1988; 64:529 11. Segal KR. Use of bioelectrical impedance analysis measurements as an evaluation for participating in sports. Am J Clin Nutr 1996;64:469S 12. Kushner RF, Kunigk A, Alspaugh M, et al. Validation of bioelectrical-impedance analysis as a measurement of change in body composition in obesity. Am J Clin Nutr 1990;52:219 13. Stolarczyk LM, Heyward VH, Van Loan MD, et al. The fatness-specific bioelectrical impedance analysis equations of Segal et al.: are they generalizable and practical? Am J Clin Nutr 1997;66:8 14. Segal KR, Van Loan M, Fitzgerald PI, Hodgdon JA, Van Itallie TB. Lean body mass estimation by bioelectrical impedance analysis: a four-site cross over validation. Am J Clin Nutr 1988;47:7 15. Micozzi MS, Albanes D, Jones DY, Chumlea WC. Correlations of body mass indices with weight, stature, and body composition in men and women in NHANES I and II. Am J Clin Nutr 1986;44:725
16. Frisancho R. New standards of weight and body composition by frame size and height for assessment of nutritional status of adults and the elderly. Am J Clin Nutr 1984;40:808 17. Macdonald SM, Reeder BA, Chen Y, Despres JP. Obesity in Canada: a descriptive analysis. Canadian Heart Health Surveys Research Group. Can Med Assoc J 1997;157:S3 18. Micozzi MS, Harris TM. Age variations in the relation of body mass indices to estimates of body fat and muscle mass. Am J Phys Anthropol 1990;81:375 19. Barlett HL, Puhl SM, Hodgson JL, Buskirk ER. Fat-free mass in relation to stature: ratios of fat-free mass to height in children, adults and elderly subjects. Am J Clin Nutr 1991;53:1112 20. Fentem PH, Mockett SJ. Physical activity and body composition: what do the national surveys reveal? Int J Obesity Rel Metab Disord 1998;22:S8 21. Pichard C, Kyle U, Gremion G, Gerbase M, Slosman D. Body composition by X-ray absorptiometry and bioelectrical impedance in elite female runners. Med Sci Sports Exerc 1997;29:1527 22. Pichard C, Kyle UG. Body composition measurements during wasting diseases. Curr Opin Clin Nutr 1998;1:357 23. Morabia A, Bernstein M, He´ritier S, Ylli A. Community-based surveillance of cardiovascular risk factors in Geneva: methods, resulting distributions, and comparisons with other populations. Prevent Med 1997;26:311 24. Paccaud F, Wietlisbach V, Rickenbach M. Evolution des maladies cardiovasculaires et des caracte´ristiques de l’alimentation: re´sultats de l’e´tude MONICA. In: Office Fe´de´ral de la Sante´ Publique, ed. Quatrie`me Rapport sur la Nutrition en Suisse. Bern: Office Fe´de´ral de la Sante´ Publique, 1998:198 25. Rissanen A, Helio¨vaara M, Aromaa A. Overweight and anthropometric changes in adulthood: a prospective study of 17000 Finns. Int J Obesity 1987;12:391 26. Rolland-Cachera MF, Cole TJ, Sempe´ M, et al. Body mass index variations: centiles from birth to 87 years. Eur J Clin Nutr 1991;45:13 27. Schutz Y, Je´quier E, Office Fe´de´ral de la L’obe´site´. In: Sante´ Publique, ed. Troisie`me Rapport sur la Nutrition en Suisse. Bern: Office Fe´de´ral de la Sante´ Publique, 1991:384 28. Must A, Dallal GE, Dietz WH. Reference data for obesity: 85th and 95th percentiles of body mass index (wt/ht2) and triceps skinfold thickness. Am J Clin Nutr 1991;53:839 29. Flegal KM, Carroll MD, Kuczarski RJ, Johnson CL. Overweight and obesity in the United States: prevalence and trends, 1960 –1994. Int J Obesity 1998;22:39 30. Forbes GB, Reina JC. Adult lean body mass declines with age: some longitudinal observations. Metabolism 1970;19:653 31. Chumlea WC, Guo SS, Zeller CM, Reo NV, Siervogel RM. Total body water data for white adults 18 to 64 years of age: the Fels Longitudinal Study. Kidney Int 1999;56:244 32. Bishop C, Phyllis E, Ritchey S. Norms for nutritionnal assessment of American adults by upper arm anthropometry. Am J Clin Nutr 1981;34:2530 33. Biasioli S, Foroni R, Petrosino L, et al. Effect of aging on the body composition of dialyzed subjects. Comparison with normal subjects. ASAIO J 1993;39:M596 34. Bartlett HL, Puhl SM, Hodgson JL, Burskirk ER. Fat-free mass in relation to stature: ratios of fat-free mass to height in children, adults, and elderly subjects. Am J Clin Nutr 1991;53:1112 35. Kehayias JJ, Fiatarone MA, Zhuang H, Roubenoff R. Total body potassium and body fat: relevance to aging. Am J Clin Nutr 1997;66:904 36. Guo SS, Zeller C, Chumlea WC, Siervogel RM. Aging, body composition, and lifestyle: the Fels Longitudinal Study. Am J Clin Nutr 1999;70:405 37. Mott JW, Wang J, Thornton JC, et al. Relation between body fat and age in 4 ethnic groups. Am J Clin Nutr 1999:1007 38. Deurenberg P, Weststrate JA, Seidell JC. Body mass index as a measure of body fatness: age- and sex-specific prediction formula. Br J Nutr 1991;65:105 39. Abernathy RP, Black DR. Healthy body weights: an alternative perspective. Am J Clin Nutr 1996;63:448S 40. de Onis M, Habicht JP. Anthropometric reference data for international use: recommendations from a World Health Organization Expert Committee. Am J Clin Nutr 1996;64:650
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
253
ANNEX 1. PHYSICAL CHARACTERISTICS (MEAN ⫾ SD) OF VALIDATION GROUPS Non-obese n
Men (80)
Women (69)
Overweight and obese Men (58)
Women (20)
Age (y) 41.2 ⫾ 13.5 43.2 ⫾ 11.7 48.2 ⫾ 10.7 53.3 ⫾ 5.2 Height (cm) 174.7 ⫾ 6.6 160.6 ⫾ 5.7 174.4 ⫾ 6.9 158.8 ⫾ 6.4 Weight (kg) 67.5 ⫾ 9.4 55.0 ⫾ 10.9 86.8 ⫾ 8.8 81.5 ⫾ 8.1 21.2 ⫾ 3.5 28.0 ⫾ 4.1 32.3 ⫾ 1.6 BMI (kg/cm2) 22.1 ⫾ 2.9 IBW (%) 96.4 ⫾ 13.0 94.4 ⫾ 16.1 124.1 ⫾ 10.3 137.9 ⫾ 4.8 FFMDXA (kg) 54.4 ⫾ 6.8 39.3 ⫾ 5.9 63.3 ⫾ 7.0 46.5 ⫾ 4.0 FFMBIA (kg) 54.5 ⫾ 7.5 39.7 ⫾ 6.0 62.9 ⫾ 6.8 45.6 ⫾ 3.9 BMI, body mass index; FFMBIA, fat-free mass measured by bioelectrical impedance analysis; FFMDXA, fat-free mass measured by dual-energy x-ray absorptiometry (DXA); IBW, ideal body weight.
ANNEX 3. Correlations (top) and differences (bottom) of FFM in nonobese women estimated by BIA and dual-energy x-ray absorptiometry (Hologic QDR-4500). The difference (calculated as FFMDXA ⫺ [FFMDXA ⫹ FFMBIA/2 per Bland-Altman]) is plotted against the mean of the measurements of FFM by DXA and BIA (FFMDXA ⫹ FFMBIA/2). Dotted line in top graph is the line of identity. BIA, bioelectrical impedance analysis; DXA, dual-energy x-ray absorptiometry; FFM, fat-free mass; SEE, standard error of the estimate.
ANNEX 2. Correlations (top) and differences (bottom) of FFM in nonobese men estimated by BIA and dual-energy x-ray absorptiometry (Hologic QDR-4500). The difference (calculated as FFMDXA ⫺ [FFMDXA ⫹ FFMBIA/2 per Bland-Altman]) is plotted against the mean of the measurements of FFM by DXA and BIA (FFMDXA ⫹ FFMBIA/2). Dotted line in top graph is the line of identity. BIA, bioelectrical impedance analysis; DXA, dual-energy x-ray absorptiometry; FFM, fat-free mass; SEE, standard error of the estimate.
254
REFERENCE VALUES OF FAT-FREE AND FAT MASSES IN HEALTHY SUBJECTS
ANNEX 4. Correlations (top) and differences (bottom) of FFM in obese men estimated by BIA and dual-energy x-ray absorptiometry (Hologic QDR-4500). The difference (calculated as FFMDXA ⫺ [FFMDXA ⫹ FFMBIA/2 per Bland-Altman]) is plotted against the mean of the measurements of FFM by DXA and BIA (FFMDXA ⫹ FFMBIA/2). Dotted line in top graph is the line of identity. BIA, bioelectrical impedance analysis; DXA, dual-energy x-ray absorptiometry; FFM, fat-free mass; SEE, standard error of the estimate.
ANNEX 5. Correlations (top) and differences (bottom) of FFM in obese women estimated by BIA and dual-energy x-ray absorptiometry (Hologic QDR-4500). The difference (calculated as FFMDXA ⫺ [FFMDXA ⫹ FFMBIA/2 per Bland-Altman]) is plotted against the mean of the measurements of FFM by DXA and BIA (FFMDXA ⫹ FFMBIA/2). Dotted line in top graph is the line of identity. BIA, bioelectrical impedance analysis; DXA, dual-energy x-ray absorptiometry; FFM, fat-free mass; SEE, standard error of the estimate.