European Journal of Clinical Nutrition (2015) 69, 837–842 © 2015 Macmillan Publishers Limited All rights reserved 0954-3007/15 www.nature.com/ejcn
ORIGINAL ARTICLE
Body composition through adult life: Swedish reference data on body composition I Larsson1, L Lissner2, G Samuelson3, H Fors4, H Lantz1, I Näslund5, LMS Carlsson6, L Sjöström6 and I Bosaeus1 BACKGROUND/OBJECTIVES: The prevalence of obesity, defined as body mass index (BMI) ⩾ 30 kg/m2, differs between populations; however, there is a need for data on description on body composition in reference populations of different ages and from different countries. The objective of this study was to pool dual-energy X-ray absorptiometry (DXA) body composition reference data from population-based Swedish cohorts. SUBJECTS/METHODS: Four population-based cross-sectional cohort studies including 1424 adult Swedes were divided into five age groups (20–29, 30–39, 40–49, 50–61 and 75 years of age); BMI 24.6 ± 3.9 kg/m2 were pooled. Body composition was measured with DXA. RESULTS: The difference in BMI from the youngest to the oldest age group was 3.2 and 4.3 kg/m2 in men and women, respectively (P o 0.001, both sexes), and fat mass (FM) was 9.9 and 9.1% higher in the oldest compared with the youngest men and women (P o 0.001, both sexes). Fat-free mass (FFM) remained stable up to 60 years of age in men (P = 0.83) and was lower at 75 years of age compared with the younger ages. In women, FFM was lower from age 60. From youngest to oldest age groups, height-adjusted FM differed from 4.6 to 7.8 kg/m2 in men and from 6.8 to 10.8 kg/m2 in women (P o0.001, both sexes). CONCLUSIONS: Our results provide reference data on body composition in Swedish populations. BMI and FM were higher among older age groups compared with the younger ones. FFM remained stable up to 60 years of age and was lower first among the 75 years of age. European Journal of Clinical Nutrition (2015) 69, 837–842; doi:10.1038/ejcn.2014.268; published online 17 December 2014
INTRODUCTION The prevalence of overweight and obesity has increased differently between countries.1 These differences depend on factors such as the composition of gender, age, ethnicity groups and several environmental and lifestyle factors within a population. Changes between fat mass (FM) depots and lean mass may also change differently between populations because of these factors. Therefore, the validity of extrapolations from one population with a specific sex-, age- and ethnic composition to another with a different composition is not known. The clinical significance of reference data of body composition includes several aspects such as understanding of changes in body composition by increasing age, in relation of sarcopenic obesity and the importance of changes in FM and fat-free mass (FFM) in relation to development of common metabolic diseases within a population. Body composition may not be superior to for example, waist circumference in relation to disease risk, but can add complementary information to this and other measurements of body fat distribution. Previously published body composition data from different populations have found secular trends of increased FM and decreased FFM during aging in longitudinal studies.2 In addition, differences in body composition between ethnic groups have been shown.2 However, some of the studies on body composition
were performed in nonrandomly selected samples3,4 and/or were limited in size.4 The aim of the present study was to compile body composition measurements by dual-energy X-ray absorptiometry (DXA) in adult men and women from four population-based studies in Sweden, in order to provide reference data for adult Swedish populations. SUBJECTS AND METHODS Four population-based studies including subjects randomly selected from the general population5–8 were pooled including, in total, 623 men and 801 women. Subjects were 20–75 years of age (mean age 45 ± 12) and a body mass index (BMI) of 24.6 ± 3.9 (16.4–45.4) kg/m2 (Table 1). The ages ranged from 27 to 61 years of age with no subjects between 22 and 26 years and 61 and 75 years. The participants were all inhabitants in SouthWestern counties of Sweden. Almost all subjects were of Swedish origin, had no serious self-reported health problems, were ambulant and were non-institutionalized. Each participant was included in only one of the four studies.
The young adult study The main aim of the young adult study was to obtain reference data for body composition in Swedish adolescents.8 The study included originally 209 boys and girls (69% of invited) aged 15 years at the first examination. Participants were re-examined at the age of 17 and 21 years, respectively.8
1 Department of Endocrinology, Diabetology and Metabolism, Internal medicine, University of Gothenburg, Sahlgrenska University Hospital, Gothenburg, Sweden; 2Public Health Epidemiology Unit, Department of Public Health and Community Medicine, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden; 3Department of Nursing, Health and Culture, University West, Trollhättan, Sweden; 4Department of Paediatrics, Northern Älvsborg Hospital, Trollhättan, Sweden; 5Department of Surgery, Örebro University Hospital, Örebro, Sweden and 6Department of Molecular and Clinical Medicine, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden. Correspondence: Dr I Larsson, Department of Endocrinology, Diabetology and Metabolism, Internal medicine, University of Gothenburg, Sahlgrenska University Hospital, Per Dubbsgatan 14, level 2, Gothenburg SE-41345, Sweden. E-mail: [email protected] Received 26 May 2014; revised 2 October 2014; accepted 18 November 2014; published online 17 December 2014
Swedish body composition reference data I Larsson et al
838 Table 1.
Age, height, weight and BMI in the pooled group and separately in the four Gothenburg population cohorts The pooled group
The young adult study8
The Mölndal Metabolic study6
The SOS reference study5
The Geriatric and Gerontologic Population Study and the Population Study of Women7
Men Numbers 623 49 50 474 50 Age, years 46.3 ± 13.5 (20–75) 20.6 ± 0.5 (20–21) 40.1 ± 15.1 (23–58) 46.4 ± 7.2 (37–61) 75 Height, cm 1.79 ± 0.06 (1.56–2.00) 1.82 ± 0.06 (1.68–1.95) 1.80 ± 0.06 (1.66–1.94) 1.79.3 ± 0.06 (1.56–2.00) 1.75 ± 0.07 (1.63–1.92)*** Weight, kg 82.3 ± 11.4 (53.0–129.0) 77.4 ± 10.1 (57.0–115.0) 81.0 ± 10.4 (62.0–109.0) 82.9 ± 11.6 (53.0–129.3) 82.0 ± 11.4 (59.1–111.0)* 25.6 ± 3.2 (17.9–37.0) 23.2 ± 2.6 (18.2–32.5) 25.0 ± 2.8 (20.1–31.1) 25.8 ± 3.2 (17.9–37.0) 26.6 ± 3.0 (19.1–32.5)*** BMI, kg/m2 BMI 25.0–29.9, % 47.0 22.4 44.0 48.9 56.0 BMI ⩾ 30, % 9.1 2.0 8.0 9.5 14.0 Women Numbers 801 Age, years 45.5 ± 12.8 (20–75) Height, cm 1.67 ± 0.07 (1.45–1.94) Weight, kg 68.3 ± 11.5 (39.5–134.3) 24.6 ± 3.9 (16.4–45.4) BMI, kg/m2 BMI 25.0–29.9, % 28.2 BMI ⩾ 30, % 9.7
56 50 20.7 ± 0.5 (20–21) 40.5 ± 15.2 (23–58) 1.67 ± 5.1 (1.57–1.77) 1.66 ± 0.06 (1.54–1.79) 63.2 ± 8.4 (48.0–83–0) 66.2 ± 9.7 (47.0–87.0) 22.8 ± 2.8 (18.3–30.1) 23.9 ± 3.4 (17.9–32.0) 14.3 22.0 1.8 8.0
647 48 45.7 ± 7.1 (37–61) 75 1.67 ± 0.07 (1.50–1.94) 1.62 ± 0.07 (1.45–1.76)*** 68.7 ± 11.5 (43.4–134.3) 70.8 ± 14.1 (39.5–119.2)** 24.6 ± 3.8 (17.6–45.4) 27.0 ± 5.0 (16.4–41.7)*** 29.5 33.3 9.3 27.1
Abbreviation: BMI, body mass index. Mean ± s.d. (range). *Po0.05, **Po 0.01, ***Po0.0001 for differences between the four study groups.
For the purpose of the present study, 49 young men and 56 young women with DXA measurements on body composition at the age of 21 years were included.
mass (TBSMM, kg) was calculated as TBSMM = 1.19 × ALST − 1.65 according to model 1 by Kim et al.10
Statistical analysis The Mölndal Metabolic study The Mölndal Metabolic study aimed to study associations between body composition, energy expenditure, dietary intake and risk factors for diabetes and cardiovascular disease in a younger and an older group of men and women.6 Fifty men and women aged 27–31 years and 50 men and women aged 57–61 years (48–68% of invited) were examined.
The Swedish Obese Subjects (SOS) reference study The SOS reference study originally included 1135 men and women in the age range of 37–61 years of age.5 In the present study, 1121 subjects with complete DXA measurements were used (Table 1). The purposes of the SOS study, which consisted of exclusively obese subjects, have been presented elsewhere.9
Descriptive anthropometry and body composition are presented as the means and s.d. for men and women separately. Relative FM was calculated as FM in kg divided by total body weight multiplied by 100. Fat mass index was calculated as FM in kg divided by height in meters squared (kg/m2), and fat-free mass index (FFMI) as well as skeletal muscle mass index were calculated as FFM and skeletal muscle mass, respectively, divided by height in meters squared (kg/m2). The subjects were divided into one of the following five age groups: 20–29, 30–39, 40–49, 50–61 and 75 years of age. Age-trend calculations were performed individually in all men and women in order to determine whether there were linear trends by age. FM and FFM were divided into percentiles by age groups in men and women, respectively. A Po0.05 significance level was used in two-sided tests. The statistical analyses were conducted with the JMP statistical software package, version 10.0.2 (2012 SAS Institute Inc., Cary, NC, USA).
The geriatric and gerontologic population study The aim of the pooled study groups of the Geriatric and Gerontologic Population Study (H70 study) was to validate bioelectrical impedance spectroscopy against DXA and establish body composition reference values among elderly Swedish men and women.7 The study was a followup of a population-based study on 70-year-old men and women, representative of their birth cohort. In the present study, 50 men and 48 women, all 75 years of age, were included. The anthropometric and DXA measurements in the four studies were performed between 1990 and 2006 in Gothenburg, Sweden. In the SOSreference study, 98 women and 79 men were measured by both anthropometry and DXA body composition in Örebro, Sweden. Body height was measured to the nearest 0.01 m with the subject standing back to a wall-mounted stadiometer in light indoor clothing with bare feet. Weight was measured to the nearest 0.1 kg with calibrated scales. BMI was calculated as weight in kg divided by height in meters squared (kg/m2). The DXA scanner used in the studies by Larsson et al.,5 Gummesson et al.6 and by Lantz et al.8 was a LUNAR DPX-L (Scanexport Medical, Helsingborg, Sweden) with software version 1.31 and with the extended analysis program for total body analysis (LUNAR Radiation, Madison, WI, USA). The technique is further described elsewhere.5 The DXA scanner used in the H70 study was a Lunar Prodigy scanner (Scanexport Medical).7 The precision of the two DXA scanners on bone mineral content and FM measurements were in the same range.5,7 Skeletal muscle mass was calculated from DXA appendicular lean soft tissue (ALST), defined as the sum of lean soft tissue in the arms and legs. Total body skeletal muscle European Journal of Clinical Nutrition (2015) 837 – 842
Ethics The ethics committee of the University of Gothenburg approved all four studies. Informed consent was obtained from all subjects before the examinations.
RESULTS Age and anthropometry In the pooled group, men were 46.3 ± 13.5 and women were 45.5 ± 12.8 years old (Table 1). BMI was 25.6 ± 3.2 and 24.6 ± 3.9 kg/m2 (range 16.4–45.4 kg/m2, both sexes). In the H70 study, both men and women were significantly shorter, heavier and had a higher BMI compared with the subjects in the other three studies. Across the five age groups from 20 to 75 years of age, BMI increased with +3.2 and +4.3 kg/m2 in men and women, respectively, Po 0.0001, both sexes (data not shown). Prevalence of overweight and obesity The prevalence of overweight and obesity defined as BMI 25.0– 29.9 and ⩾ 30 kg/m2, respectively, are presented in Table 1. The prevalence of obesity was between 1.4% in the youngest men and 14% among the oldest men (data not shown). The prevalence of © 2015 Macmillan Publishers Limited
Swedish body composition reference data I Larsson et al
839 obesity was between 2.4% in the youngest women to 27% in the oldest women.
Body composition by age groups Fat mass. DXA-determined FM increased from 20 to 75 years by 8.7 and 9.1 kg in men and women, respectively (Table 2). Correspondingly, %FM increased across age groups with 9.9 and 9.1% in men and women, respectively. Height-adjusted FM is presented in Table 2 and Figure 1a. All changes in FM across the age groups in both sexes were statistically significant, P o0.0001. FFM and skeletal muscle mass. In men and women, FFM was lower in the oldest age group (75 years) compared with the youngest (Table 3). However, when FFM in men between 20 and 60 years of age was analysed, there was no difference between the age groups (P = 0.83, not shown), whereas FFM in women 60 years of age was significantly lower compared with the younger (P o 0.05, not shown). In women, both FFMI and skeletal muscle mass index did not change statistically significantly over the age groups, whereas in men FFMI increased somewhat across the age groups to the age of 60 and became lower thereafter (Table 3, Figure 1b), while skeletal muscle mass index became lower with higher age groups in the male subjects (Table 3). In both men and women, there was a significant difference in FM between age groups in which the older subjects were found to have 9 kg higher FM compared with the youngest Although the oldest compared with the youngest men and women had lower FFM, those between 20 and 60 years of age were found to have approximately the same FFM. Percentiles of FM and FFM. At the 10th percentile, the difference between the oldest compared with youngest ages was +6.5 kg FM in men compared with +5.5 kg FM in women (Table 4). At the 90th percentile, the difference between the oldest compared with youngest ages was +4.2 kg FM in men compared with +9.8 kg FM in women (Table 4). At the 10th percentile, the oldest men had 4.5 kg less FFM than the youngest men (Table 5). The oldest women had 3 kg less FFM compared with the youngest women. At the 90th percentile, the oldest men had 4.3 kg less FFM than the younger ones. The oldest and the youngest women had
Table 2.
approximately the same amount of FFM at the 90th percentile (Table 5). DISCUSSION The usefulness of reference data on body composition such as these presented here is that reference samples obtained from different countries can be compared. In addition, from a clinical perspective, body composition per se and changes in FM and FFM by age can add complementary information to measurements of body fat distribution in relation to sarcopenic obesity and development of metabolic disease by age. In this pooled group of adult Swedish men and women, FM expressed in both absolute and relative terms increased over the age groups, whereas FFM remained stable up to ~ 60 years of age and seems to decrease from there until 75 years of age. The differences in BMI across the age groups found in the present study seem to reflect increase in FM but do not reflect the stability in FFM during middle life and the decrease in the oldest age group. Comparisons of body composition can be made with reference data from the US NHANES (National Health and Nutrition Examinations Survey), which had higher BMIs than in the present data set, and in which body composition was obtained with the same measurement technique.2 In a sample of Swiss adults where body composition was measured with bioelectrical impedance analysis consisted of healthy subjects from Western Europe.4 Another use of population-specific body composition reference data is when body fatness differs between populations at the same BMIs. Different Asian populations show higher degree of fatness compared with populations including Caucasians at the same BMI.11 Thus, population-specific body composition reference data may be useful in a discussion of different cutoffs for BMI based on body fatness and its association with disease risk.12 FFM, and specifically muscle mass, is generally considered to decrease throughout adult life.13,14 Although this seems to be true across older age groups, it may not, on a group level, be the case from young adulthood up to 61 years of age. FFMI is proposed to assess body composition in individuals who have similar body composition but differ in height15 in analogy with height adjustment of weight.16 FFMI has been proposed to allow identification of those suffering from malnutrition, wasting or those who possess a relatively high muscle mass, and also to identify sarcopenic obesity.15,17 Thus, for clinical purposes, there is
Fat mass in kg, percent and fat mass-to-height index (fat mass as kg/m2) by age groups
Age group
FM, kg
FM, %
FM index
Men 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years P for age trend All ages
15.2 ± 7.4 (3.4–44.2) 18.9 ± 6.7 (6.4–35.9) 20.0 ± 6.7 (7.0–43.1) 21.7 ± 7.4 (4.1–42.2) 23.9 ± 6.8 (9.0–37.6) o0.0001 20.1 ± 7.4 (3.4–44.2)
18.9 ± 7.2 (5.5–38.7) 22.7 ± 5.9 (11.2–36.4) 23.8 ± 5.9 (10.0–39.9) 25.2 ± 5.9 (6.9–40.9) 28.8 ± 6.3 (14.3–43.4) o 0.0001 23.9 ± 6.6 (5.5–43.4)
4.6 ± 2.2 (1.1–12.5) 5.8 ± 2.0 (2.1–11.5) 6.2 ± 2.1 (2.2–14.2) 6.8 ± 2.2 (1.3–13.5) 7.8 ± 2.3 (3.0–13.9) o 0.0001 6.3 ± 2.3 (1.1–14.2)
Women 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years P for age trend All ages
19.1 ± 6.7 (5.3–35.2) 21.1 ± 7.1 9.8–50.4) 23.1 ± 8.4 (5.1–61.5) 25.8 ± 8.1 (8.1–48.3) 28.2 ± 10.5 (6.5–63.1) o0.0001 23.4 ± 8.4 (5.1–63.1)
29.7 ± 7.0 (11.3–45.1) 31.5 ± 6.8 (17.8–55.7) 33.0 ± 7.7 (9.6–49.2) 36.4 ± 7.2 (11.4–51.9) 38.8 ± 8.1 (14.8–53.2) o 0.0001 33.7 ± 7.8 (9.6–55.7)
6.8 ± 2.5 (2.0–13.4) 7.6 ± 2.6 (3.7–20.4) 8.2 ± 3.1 (1.9–20.8) 9.5 ± 3.0 (2.5–18.0) 10.8 ± 3.9 (2.7–22.1) o 0.0001 8.5 ± 3.2 (1.9–22.1)
Abbreviations: DXA, dual-energy X-ray absorptiometry; FM, fat mass. Fat mass was measured by DXA. Mean ± s.d. (range).
© 2015 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2015) 837 – 842
Swedish body composition reference data I Larsson et al
840
Figure 1. (a) Fat mass index in men and women by age groups. Footnote to (a). Mean (95% confidence interval (CI)) men: 20–29 years (yrs): 4.6 (4.1, 5.1), 30–39 yrs: 5.8 (5.4, 6.2), 40–49 yrs: 6.2 (5.9, 6.6), 50–61 yrs: 6.8 (6.5, 7.1), 75 yrs: 7.8 (7.2, 8.4); women: 20–29 yrs: 6.8 (6.3, 7.4), 30–39 yrs: 7.6 (7.1, 8.0), 40–49 yrs: 8.2 (7.9, 8.6), 50–61 yrs: 9.5 (9.1, 9.9), 75 yrs: 10.8 (9.6, 11.9). (b) Fat-free mass index in men and women by age groups. Footnote to (b). Mean (95% CI) men: 20–29 yrs: 18.8 (18.6, 19.2), 30–39 yrs: 19.1 (18.8, 19.4), 40–49 yrs: 19.4 (19.2, 19.7), 50–61 yrs: 19.5 (19.2, 19.7), 75 yrs: 18.9 (18.4, 19.3); women: 20–29 yrs: 15.6 (15.4, 15.8), 30–39 yrs: 15.9 (15.7, 16.1), 40–49 yrs: 16.0 (15.8, 16.2), 50–61 yrs: 16.0 (15.8, 16.2), 75 yrs: 16.1 (15.7, 16.5).
Table 3.
an increasing interest to assess nutritional status in terms of height-adjusted body compartments. In women, we found that FFMI was not significantly lower, although FFM was, in the older age groups. This can be seen as contradictory but the denominator in this ratio; height did not change until the age groups of 61 years and older. Thus, with squared height, as an unchanged denominator, there were only minor changes in the ratio. In males, height decreased from the youngest age group through the four older age groups and, accordingly, male FFMI increased somewhat to the age of 61 years. This shows the importance of different heights in relation to body components. The stability in FFM up to 61 years of age found in both men and women in the present study may be explained by that fact that it is especially healthy individuals who have an active lifestyle that agree to participate in this type of study. We have no information on physical activity level in the study groups to confirm or reject that the older subjects are physically active for their age group. On the other hand, we do not know how much physical activity is enough to preserve FFM when growing older. The four studies included in the present pooled cohort are randomly selected groups from the general population and the results differ from other body composition studies that did not include randomly selected groups.3,4 Thus, the generalizability of these results should be acceptable to adults between 20 and 61 years of age. Although a lack of a continuum in the age interval, between 62 and 74 years, we think it is reasonable to assume the appropriateness that FM is higher and FFM is lower in subjects 75 years of age compared with those of 61 years of age as indicated in the present study. In 2010/11 in Sweden, the prevalence of overweight was 45% in men and 30% in women.18 The prevalence of obesity was 13% in men and 11.5% in women. From 1990–2006 to 2010/11, the proportion of Swedish adults with overweight has been fairly stable, whereas the proportion of obese men and women has increased. These national data based on self-reported height and weight show that BMI in men 16 years and older has increased from 24.4 to 25.8 kg/m2 from 1988/89 to 2010/11. Among women of same ages, an increase in BMI between 23.3 and 24.5 kg/m2 was shown.18 This represents an average increase of approximately 5 kg in men and 4 kg in women.
FFM, skeletal muscle mass, in kg, percent and indices by age groups
Age group
FFM, kg
FFM index
SM, kg
SM, %
SM index
Men 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years P for age trend All ages
62.8 ± 6.5 (47.6–79.8) 62.6 ± 6.8 (47.8–83.8) 62.4 ± 6.0 (42.2–88.8) 62.1 ± 6.9 (44.7–86.8) 58.2 ± 7.9 (44.6–82.6) o0.01 52.1 ± 23.3 (42.2–82.6)
18.9 ± 1.3 (15.5–21.7) 19.1 ± 1.5 (15.5–22.9) 19.4 ± 1.5 (15.0–23.8) 19.5 ± 1.7 (15.5–24.8) 18.9 ± 1.7 (15.2–22.6) o0.05 19.3 ± 1.6 (15.0–24.8)
33.8 ± 3.9 (25.1–44.5) 32.7 ± 3.9 (21.7–43.0) 32.7 ± 3.5 (23.3–40.6) 32.3 ± 3.8 (21.8–44.3) 29.0 ± 4.3 (21.1–39.9) o0.0001 32.4 ± 4.0 (21.1–44.5)
43.7 ± 3.9 (33.8–51.1) 40.7 ± 3.3 (31.4–47.7) 39.8 ± 3.5 (29.4–48.1) 39.1 ± 3.2 (30.3–47.5) 35.5 ± 3.5 (27.3–42.8) o 0.0001 39.8 ± 4.0 (27.3–51.1)
10.2 ± 0.8 (8.2–12.0) 10.0 ± 1.0 (7.5–12.8) 10.1 ± 0.9 (7.4–12.5) 10.1 ± 0.9 (7.4–12.9) 9.4 ± 1.0 (7.0–11.4) o 0.0001 10.0 ± 0.9 (7.0–12.9)
Women 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years P for age trend All ages
43.8 ± 3.8 (34.9–53.3) 44.8 ± 6.0 (34.1–69.1) 45.4 ± 7.6 (33.5–77–7) 43.8 ± 6.1 (31.6–71.3) 42.4 ± 5.2 (32.8–55.4) o0.01 40.2 ± 14.7 (33.5–77.7)
15.6 ± 0.8 (13.6–17.7) 15.9 ± 1.4 (12.7–21.3) 16.0 ± 1.7 (13.3–21.3) 16.0 ± 1.5 (13.0–23.0) 16.1 ± 1.5 (13.6–19.4) 0.28 15.9 ± 1.5 (12.7–23.0)
22.2 ± 2.3 (17.5–28.3) 22.4 ± 3.7 (15.7–38.9) 22.6 ± 4.6 (15.3–39.4) 21.8 ± 3.6 (15.6–37.1) 20.0 ± 2.7 (14.8–25.4) o0.001 22.1 ± 3.9 (14.8–39.4)
35.6 ± 3.6 (26.8–42.8) 34.1 ± 3.9 (21.4–45.8) 33.4 ± 4.3 (23.4–46.5) 31.6 ± 3.9 (22.7–46.5) 29.0 ± 4.2 (20.9–40.7) o 0.0001 33.0 ± 4.4 (20.9–46.5)
7.9 ± 0.6 (6.8–9.3) 7.9 ± 0.9 (5.6–11.5) 8.0 ± 1.0 (5.8–11.5) 7.9 ± 0.9 (6.3–11.4) 7.6 ± 0.7 (5.8–9.6) 0.22 7.9 ± 0.9 (5.6–11.5)
Abbreviations: DXA, dual-energy X-ray absorptiometry; FFM, fat-free mass; FM, fat mass; FFM index, FFM in kg/height m2; SM, skeletal muscle; SM index, SM in kg/height, m2. FFM was the sum of lean body mass and bone mineral content measured by DXA. Mean ± s.d. (range).
European Journal of Clinical Nutrition (2015) 837 – 842
© 2015 Macmillan Publishers Limited
Swedish body composition reference data I Larsson et al
841 Table 4.
Fat mass by percentiles in different age groups n
2.5th
10th
25th
50th
75th
90th
97.5th
Men 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years All ages
74 108 177 214 50
4.2 7.8 7.9 7.7 9.3 7.2
7.0 10.0 12.1 12.3 13.5 10.4
10.2 14.4 14.9 16.9 19.6 14.6
13.4 18.8 20.0 21.0 24.7 19.3
17.7 23.1 24.1 26.8 28.7 25.0
27.1 27.0 28.6 32.0 31.3 29.9
35.4 35.2 36.3 36.6 37.1 36.0
Women 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years All ages
82 155 274 242 48
7.9 10.9 10.7 11.7 6.9 10.2
11.5 13.4 13.8 16.7 17.0 13.7
14.1 16.4 16.9 20.2 21.3 17.2
18.0 19.5 21.5 25.1 26.7 21.5
22.4 24.7 27.8 30.7 34.9 28.0
31.7 31.4 35.0 37.1 41.5 34.3
35.1 43.9 45.2 43.7 47.1 43.4
Age group
Abbreviation: DXA, dual-energy X-ray absorptiometry. Fat mass was measured by DXA. Data shown are percentiles of fat-free mass in kg in men and women.
Table 5.
Fat-free mass by percentiles in different age groups n
2.5th
10th
25th
50th
75th
90th
97.5th
Men 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years All ages
74 108 177 214 50
50.7 49.7 49.3 50.1 45.1 48.9
52.8 54.8 54.7 53.1 48.3 53.1
58.9 57.3 59.0 57.4 51.8 57.3
63.0 61.8 62.5 61.7 58.6 61.8
68.1 67.0 66.0 66.5 62.6 66.1
71.0 72.2 70.0 71.4 66.7 70.5
74.6 76.5 73.3 75.5 80.7 74.2
Women 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years All ages
82 155 274 242 48
36.8 35.8 35.3 35.1 33.0 35.2
38.4 37.7 37.6 37.9 35.4 37.8
41.3 41.2 40.1 40.6 39.5 40.6
43.2 44.2 43.9 42.8 41.5 43.4
46.3 47.3 48.0 45.7 45.6 47.0
48.4 51.7 55.0 49.8 49.1 52.3
53.2 63.4 66.1 63.5 54.9 65.5
Age group
Abbreviation: DXA, dual-energy X-ray absorptiometry. Fat mass was measured by DXA. Data shown are percentiles of fat-free mass in kg in men and women.
In the NHANES 1999–2004,2 the prevalence of overweight was ~ 41% in men and 28% in women. Correspondingly, prevalence of obesity was ~ 28% in men and 33% in women. The NHANES populations include large groups of different ethnicities including Caucasians, with different prevalence of overweight and obesity, whereas in the present study almost all are of Scandinavian origin. The differences in the ethnic background, average BMI, prevalence of overweight and obesity between populations call for the need of more population-specific body composition reference data. Although this pooled data set from four population-based studies has a limited sample size compared with, for example, NHANES study groups,2 the presentation of FM and FFM in commonly used percentiles by age groups makes it possible to compare percentile ranges with other study groups.2 There are limitations of the present study. In a longitudinal study, true changes over time can be described. As this was not possible, the cross-sectional design of the present study has to be seen as a limitation and, thus, we cannot exclude that secular trends in weight or body composition may have influenced the results. The vast majority of the subjects (497%) included in the four studies were healthy Caucasians (Northwest European). Thus, conclusions cannot be extended to groups of other ethnicities. © 2015 Macmillan Publishers Limited
The weight limitation of the Lunar-DPX-L DXA equipment resulted in that subjects with body weights above 110 kg could not be included, and thus limited the study group regarding body weight and body composition. The anthropometric and body composition data in the four studies that constitute this study population were collected during 16 years. Over such a time period, secular trends in prevalence of overweight and obesity that may also affect body composition components that is, FFM and FM. Another limitation is that we do not have data on physical activity level of the participants. With such we may have been able to discuss the findings in FFM across age groups. One of the strength with the study is that all participants were inhabitants in the same region of Sweden. Moreover, all four of the included studies used the DXA body composition technique and DXA equipment from the same manufacturer. All anthropometric data were measured, avoiding potential bias of selfreported height and weight. When the four studies were pooled, sample size became large compared with European body composition studies4 but small relative to U.S. datasets.2 CONCLUSIONS In these reference data on body composition in Swedish populations, BMI and FM were higher among older age groups compared with the younger ones. FFM remained stable up to 60 years of age and was lower first among the 75 years of age. As the prevalence of overweight and obesity differs between populations, so also body components, in particular FM and FFM; population-specific reference data are important. The use of such information in research areas, including nutrition, energy metabolism and obesity-specific diseases, is valuable to increase the validity of conclusions regarding body composition in specific populations and may protect from spurious assumptions about body composition data that are not valid for the studied group. CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS We acknowledge Björn Henning, Institute of Medicine, Sahlgrenska Academy, University of Gothenburg for his skilful work with the DXA body composition databases.
REFERENCES 1 Ahima RS. Digging deeper into obesity. J Clin Invest 2011; 121: 2076–2079. 2 Li C, Ford ES, Zhao G, Balluz LS, Giles WH. Estimates of body composition with dual-energy X-ray absorptiometry in adults. Am J Clin Nutr 2009; 90: 1457–1465. 3 Schutz Y, Kyle UU, Pichard C. Fat-free mass index and fat mass index percentiles in Caucasians aged 18-98 y. Int J Obes Relat Metab Disord 2002; 26: 953–960. 4 Kyle UG, Melzer K, Kayser B, Picard-Kossovsky M, Gremion G, Pichard C. Eight-year longitudinal changes in body composition in healthy Swiss adults. J Am Coll Nutr 2006; 25: 493–501. 5 Larsson I, Berteus Forslund H, Lindroos AK, Lissner L, Naslund I, Peltonen M et al. Body composition in the SOS (Swedish Obese Subjects) reference study. Int J Obes Relat Metab Disord 2004; 28: 1317–1324. 6 Gummesson A, Jernas M, Svensson PA, Larsson I, Glad CA, Schele E et al. Relations of adipose tissue CIDEA gene expression to basal metabolic rate, energy restriction, and obesity: population-based and dietary intervention studies. J Clin Endocrinol Metab 2007; 92: 4759–4765. 7 Tengvall M, Ellegard L, Malmros V, Bosaeus N, Lissner L, Bosaeus I. Body composition in the elderly: reference values and bioelectrical impedance spectroscopy to predict total body skeletal muscle mass. Clin Nutr 2009; 28: 52–58. 8 Lantz H, Bratteby L-E, Fors H, Sandhagen B, Sjöström L, Samuelson G. Body composition in a cohort of Swedish adolescents aged 15, 17 and 20.5 years. Acta Pædiatrica 2008; 97: 1691–1697. 9 Sjöström L, Larsson B, Backman L, Bengtsson C, Bouchard C, Dahlgren S et al. Swedish obese subjects (SOS). Recruitment for an intervention study and a selected description of the obese state. Int J Obes Relat Metab Disord 1992; 16: 465–479.
European Journal of Clinical Nutrition (2015) 837 – 842
Swedish body composition reference data I Larsson et al
842 10 Kim J, Heshka S, Gallagher D, Kotler DP, Mayer L, Albu J et al. Intermuscular adipose tissue-free skeletal muscle mass: estimation by dual-energy X-ray absorptiometry in adults. J Appl Physiol 2004; 97: 655–660. 11 Deurenberg-Yap M, Chew SK, Deurenberg P. Elevated body fat percentage and cardiovascular risks at low body mass index levels among Singaporean Chinese, Malays and Indians. Obes Rev 2002; 3: 209–215. 12 Deurenberg-Yap M, Deurenberg P. Is a re-evaluation of WHO body composition cut-off values needed? The case of Asians in Singapore. Nutr Rev 2003; 61: S80–S87. 13 Baumgartner RN, Heymsfield SB, Roche AF. Human body composition and the epidemiology of chronic disease. Obes Res 1993; 3: 73–95. 14 Buffa R, Floris GU, Putzu PF, Marini E. Body composition variations in ageing. Collegium Antropologicum 2011; 35: 259–265.
European Journal of Clinical Nutrition (2015) 837 – 842
15 Kyle UG, Schutz Y, Dupertius YM, Pichard C. Body composition Interpretation: contribution of fet-free mass index and the body fat mass index. Nutrition 2003; 19: 597–604. 16 Larsson I, Henning B, Lindroos AK, Näslund I, Sjöström CD, Sjöström L. Optimized predictions of absolute and relative amounts of body fat from weight, height, other anthropometric predictors and age. Am J Clin Nutr 2006; 83: 252–259. 17 Hull HR, Thornton J, Wang J, Pierson RN Jr, Kaleem Z, Pi-Sunyer X et al. Fat-free mass index: changes and race/ethnic differences in adulthood. Int J Obes (Lond) 2011; 35: 121–127. 18 Samuelsson C, Hagman A. We are Putting on Weight on the Waistline. (Vi växer på bredden). Statistics Sweden: Stockholm, 2012.
© 2015 Macmillan Publishers Limited
ORIGINAL ARTICLE
Body composition through adult life: Swedish reference data on body composition I Larsson1, L Lissner2, G Samuelson3, H Fors4, H Lantz1, I Näslund5, LMS Carlsson6, L Sjöström6 and I Bosaeus1 BACKGROUND/OBJECTIVES: The prevalence of obesity, defined as body mass index (BMI) ⩾ 30 kg/m2, differs between populations; however, there is a need for data on description on body composition in reference populations of different ages and from different countries. The objective of this study was to pool dual-energy X-ray absorptiometry (DXA) body composition reference data from population-based Swedish cohorts. SUBJECTS/METHODS: Four population-based cross-sectional cohort studies including 1424 adult Swedes were divided into five age groups (20–29, 30–39, 40–49, 50–61 and 75 years of age); BMI 24.6 ± 3.9 kg/m2 were pooled. Body composition was measured with DXA. RESULTS: The difference in BMI from the youngest to the oldest age group was 3.2 and 4.3 kg/m2 in men and women, respectively (P o 0.001, both sexes), and fat mass (FM) was 9.9 and 9.1% higher in the oldest compared with the youngest men and women (P o 0.001, both sexes). Fat-free mass (FFM) remained stable up to 60 years of age in men (P = 0.83) and was lower at 75 years of age compared with the younger ages. In women, FFM was lower from age 60. From youngest to oldest age groups, height-adjusted FM differed from 4.6 to 7.8 kg/m2 in men and from 6.8 to 10.8 kg/m2 in women (P o0.001, both sexes). CONCLUSIONS: Our results provide reference data on body composition in Swedish populations. BMI and FM were higher among older age groups compared with the younger ones. FFM remained stable up to 60 years of age and was lower first among the 75 years of age. European Journal of Clinical Nutrition (2015) 69, 837–842; doi:10.1038/ejcn.2014.268; published online 17 December 2014
INTRODUCTION The prevalence of overweight and obesity has increased differently between countries.1 These differences depend on factors such as the composition of gender, age, ethnicity groups and several environmental and lifestyle factors within a population. Changes between fat mass (FM) depots and lean mass may also change differently between populations because of these factors. Therefore, the validity of extrapolations from one population with a specific sex-, age- and ethnic composition to another with a different composition is not known. The clinical significance of reference data of body composition includes several aspects such as understanding of changes in body composition by increasing age, in relation of sarcopenic obesity and the importance of changes in FM and fat-free mass (FFM) in relation to development of common metabolic diseases within a population. Body composition may not be superior to for example, waist circumference in relation to disease risk, but can add complementary information to this and other measurements of body fat distribution. Previously published body composition data from different populations have found secular trends of increased FM and decreased FFM during aging in longitudinal studies.2 In addition, differences in body composition between ethnic groups have been shown.2 However, some of the studies on body composition
were performed in nonrandomly selected samples3,4 and/or were limited in size.4 The aim of the present study was to compile body composition measurements by dual-energy X-ray absorptiometry (DXA) in adult men and women from four population-based studies in Sweden, in order to provide reference data for adult Swedish populations. SUBJECTS AND METHODS Four population-based studies including subjects randomly selected from the general population5–8 were pooled including, in total, 623 men and 801 women. Subjects were 20–75 years of age (mean age 45 ± 12) and a body mass index (BMI) of 24.6 ± 3.9 (16.4–45.4) kg/m2 (Table 1). The ages ranged from 27 to 61 years of age with no subjects between 22 and 26 years and 61 and 75 years. The participants were all inhabitants in SouthWestern counties of Sweden. Almost all subjects were of Swedish origin, had no serious self-reported health problems, were ambulant and were non-institutionalized. Each participant was included in only one of the four studies.
The young adult study The main aim of the young adult study was to obtain reference data for body composition in Swedish adolescents.8 The study included originally 209 boys and girls (69% of invited) aged 15 years at the first examination. Participants were re-examined at the age of 17 and 21 years, respectively.8
1 Department of Endocrinology, Diabetology and Metabolism, Internal medicine, University of Gothenburg, Sahlgrenska University Hospital, Gothenburg, Sweden; 2Public Health Epidemiology Unit, Department of Public Health and Community Medicine, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden; 3Department of Nursing, Health and Culture, University West, Trollhättan, Sweden; 4Department of Paediatrics, Northern Älvsborg Hospital, Trollhättan, Sweden; 5Department of Surgery, Örebro University Hospital, Örebro, Sweden and 6Department of Molecular and Clinical Medicine, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden. Correspondence: Dr I Larsson, Department of Endocrinology, Diabetology and Metabolism, Internal medicine, University of Gothenburg, Sahlgrenska University Hospital, Per Dubbsgatan 14, level 2, Gothenburg SE-41345, Sweden. E-mail: [email protected] Received 26 May 2014; revised 2 October 2014; accepted 18 November 2014; published online 17 December 2014
Swedish body composition reference data I Larsson et al
838 Table 1.
Age, height, weight and BMI in the pooled group and separately in the four Gothenburg population cohorts The pooled group
The young adult study8
The Mölndal Metabolic study6
The SOS reference study5
The Geriatric and Gerontologic Population Study and the Population Study of Women7
Men Numbers 623 49 50 474 50 Age, years 46.3 ± 13.5 (20–75) 20.6 ± 0.5 (20–21) 40.1 ± 15.1 (23–58) 46.4 ± 7.2 (37–61) 75 Height, cm 1.79 ± 0.06 (1.56–2.00) 1.82 ± 0.06 (1.68–1.95) 1.80 ± 0.06 (1.66–1.94) 1.79.3 ± 0.06 (1.56–2.00) 1.75 ± 0.07 (1.63–1.92)*** Weight, kg 82.3 ± 11.4 (53.0–129.0) 77.4 ± 10.1 (57.0–115.0) 81.0 ± 10.4 (62.0–109.0) 82.9 ± 11.6 (53.0–129.3) 82.0 ± 11.4 (59.1–111.0)* 25.6 ± 3.2 (17.9–37.0) 23.2 ± 2.6 (18.2–32.5) 25.0 ± 2.8 (20.1–31.1) 25.8 ± 3.2 (17.9–37.0) 26.6 ± 3.0 (19.1–32.5)*** BMI, kg/m2 BMI 25.0–29.9, % 47.0 22.4 44.0 48.9 56.0 BMI ⩾ 30, % 9.1 2.0 8.0 9.5 14.0 Women Numbers 801 Age, years 45.5 ± 12.8 (20–75) Height, cm 1.67 ± 0.07 (1.45–1.94) Weight, kg 68.3 ± 11.5 (39.5–134.3) 24.6 ± 3.9 (16.4–45.4) BMI, kg/m2 BMI 25.0–29.9, % 28.2 BMI ⩾ 30, % 9.7
56 50 20.7 ± 0.5 (20–21) 40.5 ± 15.2 (23–58) 1.67 ± 5.1 (1.57–1.77) 1.66 ± 0.06 (1.54–1.79) 63.2 ± 8.4 (48.0–83–0) 66.2 ± 9.7 (47.0–87.0) 22.8 ± 2.8 (18.3–30.1) 23.9 ± 3.4 (17.9–32.0) 14.3 22.0 1.8 8.0
647 48 45.7 ± 7.1 (37–61) 75 1.67 ± 0.07 (1.50–1.94) 1.62 ± 0.07 (1.45–1.76)*** 68.7 ± 11.5 (43.4–134.3) 70.8 ± 14.1 (39.5–119.2)** 24.6 ± 3.8 (17.6–45.4) 27.0 ± 5.0 (16.4–41.7)*** 29.5 33.3 9.3 27.1
Abbreviation: BMI, body mass index. Mean ± s.d. (range). *Po0.05, **Po 0.01, ***Po0.0001 for differences between the four study groups.
For the purpose of the present study, 49 young men and 56 young women with DXA measurements on body composition at the age of 21 years were included.
mass (TBSMM, kg) was calculated as TBSMM = 1.19 × ALST − 1.65 according to model 1 by Kim et al.10
Statistical analysis The Mölndal Metabolic study The Mölndal Metabolic study aimed to study associations between body composition, energy expenditure, dietary intake and risk factors for diabetes and cardiovascular disease in a younger and an older group of men and women.6 Fifty men and women aged 27–31 years and 50 men and women aged 57–61 years (48–68% of invited) were examined.
The Swedish Obese Subjects (SOS) reference study The SOS reference study originally included 1135 men and women in the age range of 37–61 years of age.5 In the present study, 1121 subjects with complete DXA measurements were used (Table 1). The purposes of the SOS study, which consisted of exclusively obese subjects, have been presented elsewhere.9
Descriptive anthropometry and body composition are presented as the means and s.d. for men and women separately. Relative FM was calculated as FM in kg divided by total body weight multiplied by 100. Fat mass index was calculated as FM in kg divided by height in meters squared (kg/m2), and fat-free mass index (FFMI) as well as skeletal muscle mass index were calculated as FFM and skeletal muscle mass, respectively, divided by height in meters squared (kg/m2). The subjects were divided into one of the following five age groups: 20–29, 30–39, 40–49, 50–61 and 75 years of age. Age-trend calculations were performed individually in all men and women in order to determine whether there were linear trends by age. FM and FFM were divided into percentiles by age groups in men and women, respectively. A Po0.05 significance level was used in two-sided tests. The statistical analyses were conducted with the JMP statistical software package, version 10.0.2 (2012 SAS Institute Inc., Cary, NC, USA).
The geriatric and gerontologic population study The aim of the pooled study groups of the Geriatric and Gerontologic Population Study (H70 study) was to validate bioelectrical impedance spectroscopy against DXA and establish body composition reference values among elderly Swedish men and women.7 The study was a followup of a population-based study on 70-year-old men and women, representative of their birth cohort. In the present study, 50 men and 48 women, all 75 years of age, were included. The anthropometric and DXA measurements in the four studies were performed between 1990 and 2006 in Gothenburg, Sweden. In the SOSreference study, 98 women and 79 men were measured by both anthropometry and DXA body composition in Örebro, Sweden. Body height was measured to the nearest 0.01 m with the subject standing back to a wall-mounted stadiometer in light indoor clothing with bare feet. Weight was measured to the nearest 0.1 kg with calibrated scales. BMI was calculated as weight in kg divided by height in meters squared (kg/m2). The DXA scanner used in the studies by Larsson et al.,5 Gummesson et al.6 and by Lantz et al.8 was a LUNAR DPX-L (Scanexport Medical, Helsingborg, Sweden) with software version 1.31 and with the extended analysis program for total body analysis (LUNAR Radiation, Madison, WI, USA). The technique is further described elsewhere.5 The DXA scanner used in the H70 study was a Lunar Prodigy scanner (Scanexport Medical).7 The precision of the two DXA scanners on bone mineral content and FM measurements were in the same range.5,7 Skeletal muscle mass was calculated from DXA appendicular lean soft tissue (ALST), defined as the sum of lean soft tissue in the arms and legs. Total body skeletal muscle European Journal of Clinical Nutrition (2015) 837 – 842
Ethics The ethics committee of the University of Gothenburg approved all four studies. Informed consent was obtained from all subjects before the examinations.
RESULTS Age and anthropometry In the pooled group, men were 46.3 ± 13.5 and women were 45.5 ± 12.8 years old (Table 1). BMI was 25.6 ± 3.2 and 24.6 ± 3.9 kg/m2 (range 16.4–45.4 kg/m2, both sexes). In the H70 study, both men and women were significantly shorter, heavier and had a higher BMI compared with the subjects in the other three studies. Across the five age groups from 20 to 75 years of age, BMI increased with +3.2 and +4.3 kg/m2 in men and women, respectively, Po 0.0001, both sexes (data not shown). Prevalence of overweight and obesity The prevalence of overweight and obesity defined as BMI 25.0– 29.9 and ⩾ 30 kg/m2, respectively, are presented in Table 1. The prevalence of obesity was between 1.4% in the youngest men and 14% among the oldest men (data not shown). The prevalence of © 2015 Macmillan Publishers Limited
Swedish body composition reference data I Larsson et al
839 obesity was between 2.4% in the youngest women to 27% in the oldest women.
Body composition by age groups Fat mass. DXA-determined FM increased from 20 to 75 years by 8.7 and 9.1 kg in men and women, respectively (Table 2). Correspondingly, %FM increased across age groups with 9.9 and 9.1% in men and women, respectively. Height-adjusted FM is presented in Table 2 and Figure 1a. All changes in FM across the age groups in both sexes were statistically significant, P o0.0001. FFM and skeletal muscle mass. In men and women, FFM was lower in the oldest age group (75 years) compared with the youngest (Table 3). However, when FFM in men between 20 and 60 years of age was analysed, there was no difference between the age groups (P = 0.83, not shown), whereas FFM in women 60 years of age was significantly lower compared with the younger (P o 0.05, not shown). In women, both FFMI and skeletal muscle mass index did not change statistically significantly over the age groups, whereas in men FFMI increased somewhat across the age groups to the age of 60 and became lower thereafter (Table 3, Figure 1b), while skeletal muscle mass index became lower with higher age groups in the male subjects (Table 3). In both men and women, there was a significant difference in FM between age groups in which the older subjects were found to have 9 kg higher FM compared with the youngest Although the oldest compared with the youngest men and women had lower FFM, those between 20 and 60 years of age were found to have approximately the same FFM. Percentiles of FM and FFM. At the 10th percentile, the difference between the oldest compared with youngest ages was +6.5 kg FM in men compared with +5.5 kg FM in women (Table 4). At the 90th percentile, the difference between the oldest compared with youngest ages was +4.2 kg FM in men compared with +9.8 kg FM in women (Table 4). At the 10th percentile, the oldest men had 4.5 kg less FFM than the youngest men (Table 5). The oldest women had 3 kg less FFM compared with the youngest women. At the 90th percentile, the oldest men had 4.3 kg less FFM than the younger ones. The oldest and the youngest women had
Table 2.
approximately the same amount of FFM at the 90th percentile (Table 5). DISCUSSION The usefulness of reference data on body composition such as these presented here is that reference samples obtained from different countries can be compared. In addition, from a clinical perspective, body composition per se and changes in FM and FFM by age can add complementary information to measurements of body fat distribution in relation to sarcopenic obesity and development of metabolic disease by age. In this pooled group of adult Swedish men and women, FM expressed in both absolute and relative terms increased over the age groups, whereas FFM remained stable up to ~ 60 years of age and seems to decrease from there until 75 years of age. The differences in BMI across the age groups found in the present study seem to reflect increase in FM but do not reflect the stability in FFM during middle life and the decrease in the oldest age group. Comparisons of body composition can be made with reference data from the US NHANES (National Health and Nutrition Examinations Survey), which had higher BMIs than in the present data set, and in which body composition was obtained with the same measurement technique.2 In a sample of Swiss adults where body composition was measured with bioelectrical impedance analysis consisted of healthy subjects from Western Europe.4 Another use of population-specific body composition reference data is when body fatness differs between populations at the same BMIs. Different Asian populations show higher degree of fatness compared with populations including Caucasians at the same BMI.11 Thus, population-specific body composition reference data may be useful in a discussion of different cutoffs for BMI based on body fatness and its association with disease risk.12 FFM, and specifically muscle mass, is generally considered to decrease throughout adult life.13,14 Although this seems to be true across older age groups, it may not, on a group level, be the case from young adulthood up to 61 years of age. FFMI is proposed to assess body composition in individuals who have similar body composition but differ in height15 in analogy with height adjustment of weight.16 FFMI has been proposed to allow identification of those suffering from malnutrition, wasting or those who possess a relatively high muscle mass, and also to identify sarcopenic obesity.15,17 Thus, for clinical purposes, there is
Fat mass in kg, percent and fat mass-to-height index (fat mass as kg/m2) by age groups
Age group
FM, kg
FM, %
FM index
Men 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years P for age trend All ages
15.2 ± 7.4 (3.4–44.2) 18.9 ± 6.7 (6.4–35.9) 20.0 ± 6.7 (7.0–43.1) 21.7 ± 7.4 (4.1–42.2) 23.9 ± 6.8 (9.0–37.6) o0.0001 20.1 ± 7.4 (3.4–44.2)
18.9 ± 7.2 (5.5–38.7) 22.7 ± 5.9 (11.2–36.4) 23.8 ± 5.9 (10.0–39.9) 25.2 ± 5.9 (6.9–40.9) 28.8 ± 6.3 (14.3–43.4) o 0.0001 23.9 ± 6.6 (5.5–43.4)
4.6 ± 2.2 (1.1–12.5) 5.8 ± 2.0 (2.1–11.5) 6.2 ± 2.1 (2.2–14.2) 6.8 ± 2.2 (1.3–13.5) 7.8 ± 2.3 (3.0–13.9) o 0.0001 6.3 ± 2.3 (1.1–14.2)
Women 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years P for age trend All ages
19.1 ± 6.7 (5.3–35.2) 21.1 ± 7.1 9.8–50.4) 23.1 ± 8.4 (5.1–61.5) 25.8 ± 8.1 (8.1–48.3) 28.2 ± 10.5 (6.5–63.1) o0.0001 23.4 ± 8.4 (5.1–63.1)
29.7 ± 7.0 (11.3–45.1) 31.5 ± 6.8 (17.8–55.7) 33.0 ± 7.7 (9.6–49.2) 36.4 ± 7.2 (11.4–51.9) 38.8 ± 8.1 (14.8–53.2) o 0.0001 33.7 ± 7.8 (9.6–55.7)
6.8 ± 2.5 (2.0–13.4) 7.6 ± 2.6 (3.7–20.4) 8.2 ± 3.1 (1.9–20.8) 9.5 ± 3.0 (2.5–18.0) 10.8 ± 3.9 (2.7–22.1) o 0.0001 8.5 ± 3.2 (1.9–22.1)
Abbreviations: DXA, dual-energy X-ray absorptiometry; FM, fat mass. Fat mass was measured by DXA. Mean ± s.d. (range).
© 2015 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2015) 837 – 842
Swedish body composition reference data I Larsson et al
840
Figure 1. (a) Fat mass index in men and women by age groups. Footnote to (a). Mean (95% confidence interval (CI)) men: 20–29 years (yrs): 4.6 (4.1, 5.1), 30–39 yrs: 5.8 (5.4, 6.2), 40–49 yrs: 6.2 (5.9, 6.6), 50–61 yrs: 6.8 (6.5, 7.1), 75 yrs: 7.8 (7.2, 8.4); women: 20–29 yrs: 6.8 (6.3, 7.4), 30–39 yrs: 7.6 (7.1, 8.0), 40–49 yrs: 8.2 (7.9, 8.6), 50–61 yrs: 9.5 (9.1, 9.9), 75 yrs: 10.8 (9.6, 11.9). (b) Fat-free mass index in men and women by age groups. Footnote to (b). Mean (95% CI) men: 20–29 yrs: 18.8 (18.6, 19.2), 30–39 yrs: 19.1 (18.8, 19.4), 40–49 yrs: 19.4 (19.2, 19.7), 50–61 yrs: 19.5 (19.2, 19.7), 75 yrs: 18.9 (18.4, 19.3); women: 20–29 yrs: 15.6 (15.4, 15.8), 30–39 yrs: 15.9 (15.7, 16.1), 40–49 yrs: 16.0 (15.8, 16.2), 50–61 yrs: 16.0 (15.8, 16.2), 75 yrs: 16.1 (15.7, 16.5).
Table 3.
an increasing interest to assess nutritional status in terms of height-adjusted body compartments. In women, we found that FFMI was not significantly lower, although FFM was, in the older age groups. This can be seen as contradictory but the denominator in this ratio; height did not change until the age groups of 61 years and older. Thus, with squared height, as an unchanged denominator, there were only minor changes in the ratio. In males, height decreased from the youngest age group through the four older age groups and, accordingly, male FFMI increased somewhat to the age of 61 years. This shows the importance of different heights in relation to body components. The stability in FFM up to 61 years of age found in both men and women in the present study may be explained by that fact that it is especially healthy individuals who have an active lifestyle that agree to participate in this type of study. We have no information on physical activity level in the study groups to confirm or reject that the older subjects are physically active for their age group. On the other hand, we do not know how much physical activity is enough to preserve FFM when growing older. The four studies included in the present pooled cohort are randomly selected groups from the general population and the results differ from other body composition studies that did not include randomly selected groups.3,4 Thus, the generalizability of these results should be acceptable to adults between 20 and 61 years of age. Although a lack of a continuum in the age interval, between 62 and 74 years, we think it is reasonable to assume the appropriateness that FM is higher and FFM is lower in subjects 75 years of age compared with those of 61 years of age as indicated in the present study. In 2010/11 in Sweden, the prevalence of overweight was 45% in men and 30% in women.18 The prevalence of obesity was 13% in men and 11.5% in women. From 1990–2006 to 2010/11, the proportion of Swedish adults with overweight has been fairly stable, whereas the proportion of obese men and women has increased. These national data based on self-reported height and weight show that BMI in men 16 years and older has increased from 24.4 to 25.8 kg/m2 from 1988/89 to 2010/11. Among women of same ages, an increase in BMI between 23.3 and 24.5 kg/m2 was shown.18 This represents an average increase of approximately 5 kg in men and 4 kg in women.
FFM, skeletal muscle mass, in kg, percent and indices by age groups
Age group
FFM, kg
FFM index
SM, kg
SM, %
SM index
Men 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years P for age trend All ages
62.8 ± 6.5 (47.6–79.8) 62.6 ± 6.8 (47.8–83.8) 62.4 ± 6.0 (42.2–88.8) 62.1 ± 6.9 (44.7–86.8) 58.2 ± 7.9 (44.6–82.6) o0.01 52.1 ± 23.3 (42.2–82.6)
18.9 ± 1.3 (15.5–21.7) 19.1 ± 1.5 (15.5–22.9) 19.4 ± 1.5 (15.0–23.8) 19.5 ± 1.7 (15.5–24.8) 18.9 ± 1.7 (15.2–22.6) o0.05 19.3 ± 1.6 (15.0–24.8)
33.8 ± 3.9 (25.1–44.5) 32.7 ± 3.9 (21.7–43.0) 32.7 ± 3.5 (23.3–40.6) 32.3 ± 3.8 (21.8–44.3) 29.0 ± 4.3 (21.1–39.9) o0.0001 32.4 ± 4.0 (21.1–44.5)
43.7 ± 3.9 (33.8–51.1) 40.7 ± 3.3 (31.4–47.7) 39.8 ± 3.5 (29.4–48.1) 39.1 ± 3.2 (30.3–47.5) 35.5 ± 3.5 (27.3–42.8) o 0.0001 39.8 ± 4.0 (27.3–51.1)
10.2 ± 0.8 (8.2–12.0) 10.0 ± 1.0 (7.5–12.8) 10.1 ± 0.9 (7.4–12.5) 10.1 ± 0.9 (7.4–12.9) 9.4 ± 1.0 (7.0–11.4) o 0.0001 10.0 ± 0.9 (7.0–12.9)
Women 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years P for age trend All ages
43.8 ± 3.8 (34.9–53.3) 44.8 ± 6.0 (34.1–69.1) 45.4 ± 7.6 (33.5–77–7) 43.8 ± 6.1 (31.6–71.3) 42.4 ± 5.2 (32.8–55.4) o0.01 40.2 ± 14.7 (33.5–77.7)
15.6 ± 0.8 (13.6–17.7) 15.9 ± 1.4 (12.7–21.3) 16.0 ± 1.7 (13.3–21.3) 16.0 ± 1.5 (13.0–23.0) 16.1 ± 1.5 (13.6–19.4) 0.28 15.9 ± 1.5 (12.7–23.0)
22.2 ± 2.3 (17.5–28.3) 22.4 ± 3.7 (15.7–38.9) 22.6 ± 4.6 (15.3–39.4) 21.8 ± 3.6 (15.6–37.1) 20.0 ± 2.7 (14.8–25.4) o0.001 22.1 ± 3.9 (14.8–39.4)
35.6 ± 3.6 (26.8–42.8) 34.1 ± 3.9 (21.4–45.8) 33.4 ± 4.3 (23.4–46.5) 31.6 ± 3.9 (22.7–46.5) 29.0 ± 4.2 (20.9–40.7) o 0.0001 33.0 ± 4.4 (20.9–46.5)
7.9 ± 0.6 (6.8–9.3) 7.9 ± 0.9 (5.6–11.5) 8.0 ± 1.0 (5.8–11.5) 7.9 ± 0.9 (6.3–11.4) 7.6 ± 0.7 (5.8–9.6) 0.22 7.9 ± 0.9 (5.6–11.5)
Abbreviations: DXA, dual-energy X-ray absorptiometry; FFM, fat-free mass; FM, fat mass; FFM index, FFM in kg/height m2; SM, skeletal muscle; SM index, SM in kg/height, m2. FFM was the sum of lean body mass and bone mineral content measured by DXA. Mean ± s.d. (range).
European Journal of Clinical Nutrition (2015) 837 – 842
© 2015 Macmillan Publishers Limited
Swedish body composition reference data I Larsson et al
841 Table 4.
Fat mass by percentiles in different age groups n
2.5th
10th
25th
50th
75th
90th
97.5th
Men 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years All ages
74 108 177 214 50
4.2 7.8 7.9 7.7 9.3 7.2
7.0 10.0 12.1 12.3 13.5 10.4
10.2 14.4 14.9 16.9 19.6 14.6
13.4 18.8 20.0 21.0 24.7 19.3
17.7 23.1 24.1 26.8 28.7 25.0
27.1 27.0 28.6 32.0 31.3 29.9
35.4 35.2 36.3 36.6 37.1 36.0
Women 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years All ages
82 155 274 242 48
7.9 10.9 10.7 11.7 6.9 10.2
11.5 13.4 13.8 16.7 17.0 13.7
14.1 16.4 16.9 20.2 21.3 17.2
18.0 19.5 21.5 25.1 26.7 21.5
22.4 24.7 27.8 30.7 34.9 28.0
31.7 31.4 35.0 37.1 41.5 34.3
35.1 43.9 45.2 43.7 47.1 43.4
Age group
Abbreviation: DXA, dual-energy X-ray absorptiometry. Fat mass was measured by DXA. Data shown are percentiles of fat-free mass in kg in men and women.
Table 5.
Fat-free mass by percentiles in different age groups n
2.5th
10th
25th
50th
75th
90th
97.5th
Men 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years All ages
74 108 177 214 50
50.7 49.7 49.3 50.1 45.1 48.9
52.8 54.8 54.7 53.1 48.3 53.1
58.9 57.3 59.0 57.4 51.8 57.3
63.0 61.8 62.5 61.7 58.6 61.8
68.1 67.0 66.0 66.5 62.6 66.1
71.0 72.2 70.0 71.4 66.7 70.5
74.6 76.5 73.3 75.5 80.7 74.2
Women 20–29 Years 30–39 Years 40–49 Years 50–61 Years 75 Years All ages
82 155 274 242 48
36.8 35.8 35.3 35.1 33.0 35.2
38.4 37.7 37.6 37.9 35.4 37.8
41.3 41.2 40.1 40.6 39.5 40.6
43.2 44.2 43.9 42.8 41.5 43.4
46.3 47.3 48.0 45.7 45.6 47.0
48.4 51.7 55.0 49.8 49.1 52.3
53.2 63.4 66.1 63.5 54.9 65.5
Age group
Abbreviation: DXA, dual-energy X-ray absorptiometry. Fat mass was measured by DXA. Data shown are percentiles of fat-free mass in kg in men and women.
In the NHANES 1999–2004,2 the prevalence of overweight was ~ 41% in men and 28% in women. Correspondingly, prevalence of obesity was ~ 28% in men and 33% in women. The NHANES populations include large groups of different ethnicities including Caucasians, with different prevalence of overweight and obesity, whereas in the present study almost all are of Scandinavian origin. The differences in the ethnic background, average BMI, prevalence of overweight and obesity between populations call for the need of more population-specific body composition reference data. Although this pooled data set from four population-based studies has a limited sample size compared with, for example, NHANES study groups,2 the presentation of FM and FFM in commonly used percentiles by age groups makes it possible to compare percentile ranges with other study groups.2 There are limitations of the present study. In a longitudinal study, true changes over time can be described. As this was not possible, the cross-sectional design of the present study has to be seen as a limitation and, thus, we cannot exclude that secular trends in weight or body composition may have influenced the results. The vast majority of the subjects (497%) included in the four studies were healthy Caucasians (Northwest European). Thus, conclusions cannot be extended to groups of other ethnicities. © 2015 Macmillan Publishers Limited
The weight limitation of the Lunar-DPX-L DXA equipment resulted in that subjects with body weights above 110 kg could not be included, and thus limited the study group regarding body weight and body composition. The anthropometric and body composition data in the four studies that constitute this study population were collected during 16 years. Over such a time period, secular trends in prevalence of overweight and obesity that may also affect body composition components that is, FFM and FM. Another limitation is that we do not have data on physical activity level of the participants. With such we may have been able to discuss the findings in FFM across age groups. One of the strength with the study is that all participants were inhabitants in the same region of Sweden. Moreover, all four of the included studies used the DXA body composition technique and DXA equipment from the same manufacturer. All anthropometric data were measured, avoiding potential bias of selfreported height and weight. When the four studies were pooled, sample size became large compared with European body composition studies4 but small relative to U.S. datasets.2 CONCLUSIONS In these reference data on body composition in Swedish populations, BMI and FM were higher among older age groups compared with the younger ones. FFM remained stable up to 60 years of age and was lower first among the 75 years of age. As the prevalence of overweight and obesity differs between populations, so also body components, in particular FM and FFM; population-specific reference data are important. The use of such information in research areas, including nutrition, energy metabolism and obesity-specific diseases, is valuable to increase the validity of conclusions regarding body composition in specific populations and may protect from spurious assumptions about body composition data that are not valid for the studied group. CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS We acknowledge Björn Henning, Institute of Medicine, Sahlgrenska Academy, University of Gothenburg for his skilful work with the DXA body composition databases.
REFERENCES 1 Ahima RS. Digging deeper into obesity. J Clin Invest 2011; 121: 2076–2079. 2 Li C, Ford ES, Zhao G, Balluz LS, Giles WH. Estimates of body composition with dual-energy X-ray absorptiometry in adults. Am J Clin Nutr 2009; 90: 1457–1465. 3 Schutz Y, Kyle UU, Pichard C. Fat-free mass index and fat mass index percentiles in Caucasians aged 18-98 y. Int J Obes Relat Metab Disord 2002; 26: 953–960. 4 Kyle UG, Melzer K, Kayser B, Picard-Kossovsky M, Gremion G, Pichard C. Eight-year longitudinal changes in body composition in healthy Swiss adults. J Am Coll Nutr 2006; 25: 493–501. 5 Larsson I, Berteus Forslund H, Lindroos AK, Lissner L, Naslund I, Peltonen M et al. Body composition in the SOS (Swedish Obese Subjects) reference study. Int J Obes Relat Metab Disord 2004; 28: 1317–1324. 6 Gummesson A, Jernas M, Svensson PA, Larsson I, Glad CA, Schele E et al. Relations of adipose tissue CIDEA gene expression to basal metabolic rate, energy restriction, and obesity: population-based and dietary intervention studies. J Clin Endocrinol Metab 2007; 92: 4759–4765. 7 Tengvall M, Ellegard L, Malmros V, Bosaeus N, Lissner L, Bosaeus I. Body composition in the elderly: reference values and bioelectrical impedance spectroscopy to predict total body skeletal muscle mass. Clin Nutr 2009; 28: 52–58. 8 Lantz H, Bratteby L-E, Fors H, Sandhagen B, Sjöström L, Samuelson G. Body composition in a cohort of Swedish adolescents aged 15, 17 and 20.5 years. Acta Pædiatrica 2008; 97: 1691–1697. 9 Sjöström L, Larsson B, Backman L, Bengtsson C, Bouchard C, Dahlgren S et al. Swedish obese subjects (SOS). Recruitment for an intervention study and a selected description of the obese state. Int J Obes Relat Metab Disord 1992; 16: 465–479.
European Journal of Clinical Nutrition (2015) 837 – 842
Swedish body composition reference data I Larsson et al
842 10 Kim J, Heshka S, Gallagher D, Kotler DP, Mayer L, Albu J et al. Intermuscular adipose tissue-free skeletal muscle mass: estimation by dual-energy X-ray absorptiometry in adults. J Appl Physiol 2004; 97: 655–660. 11 Deurenberg-Yap M, Chew SK, Deurenberg P. Elevated body fat percentage and cardiovascular risks at low body mass index levels among Singaporean Chinese, Malays and Indians. Obes Rev 2002; 3: 209–215. 12 Deurenberg-Yap M, Deurenberg P. Is a re-evaluation of WHO body composition cut-off values needed? The case of Asians in Singapore. Nutr Rev 2003; 61: S80–S87. 13 Baumgartner RN, Heymsfield SB, Roche AF. Human body composition and the epidemiology of chronic disease. Obes Res 1993; 3: 73–95. 14 Buffa R, Floris GU, Putzu PF, Marini E. Body composition variations in ageing. Collegium Antropologicum 2011; 35: 259–265.
European Journal of Clinical Nutrition (2015) 837 – 842
15 Kyle UG, Schutz Y, Dupertius YM, Pichard C. Body composition Interpretation: contribution of fet-free mass index and the body fat mass index. Nutrition 2003; 19: 597–604. 16 Larsson I, Henning B, Lindroos AK, Näslund I, Sjöström CD, Sjöström L. Optimized predictions of absolute and relative amounts of body fat from weight, height, other anthropometric predictors and age. Am J Clin Nutr 2006; 83: 252–259. 17 Hull HR, Thornton J, Wang J, Pierson RN Jr, Kaleem Z, Pi-Sunyer X et al. Fat-free mass index: changes and race/ethnic differences in adulthood. Int J Obes (Lond) 2011; 35: 121–127. 18 Samuelsson C, Hagman A. We are Putting on Weight on the Waistline. (Vi växer på bredden). Statistics Sweden: Stockholm, 2012.
© 2015 Macmillan Publishers Limited
