Study design
This is a cross-sectional analysis performed in an outpatient clinic, in Liège, Belgium. Data collection was conducted from December 2013 until March 2014. The study was approved by the Ethics Committee of the University Hospital of Liège.
Subjects
Volunteers, adult subjects, were recruited from the community using advertisements in local newspapers and public places. Exclusion criteria, were the contraindications of the device: presence of an electronic implant (heart pacemaker, brain stimulator), body mass index over 50 kg/m2, limb amputation, pregnancy. All subjects gave a written informed consent prior to inclusion in the study.
To ensure adequate power, the calculation of the sample size was made before the beginning of the study. The formula used for this calculation is as follows:
$$ \mathrm{n}=\frac{2\sigma 2\left[QG\left(1-\frac{\alpha }{2}\right)+QG\left(1-\beta \right)\right]2}{\varDelta 2} $$
where, σ 2 is the variance difference between the result of appendicular lean mass obtained by BIA and the result obtained by DXA in a pilot study (=3.252 kg); α is the type I error = 0.05; β is the type II error = 0.1; Δ 2 is the maximum tolerated difference between the results obtained by BIA and by DXA (this was fixed arbitrarily to 1.27 kg, on the basis of the definition of sarcopenia by the European Working Group on Sarcopenia in Older People [11]).
The result of this calculation was n = 138, this meant that a minimum of 138 subjects were required to demonstrate any difference in muscle mass measured by the 2 studied devices. Since the Belgian population (Eurostat 2012) consists of 2.135.600 people aged 18–34 years, 4.504.729 people aged 35–64 years and 1.924.934 people aged 65 years or more, our study had to include respectively 34, 73 and 31 people in each age group in order to respect the age distribution of the population.
Anthropometric measurements
Height was measured to the nearest 0.1 cm using a stadiometer and weight to the nearest 0.1 kg using a weighting-scale. Abdominal circumference was also measured, to the nearest 0.1 cm, using a tape measure at the navel. Muscle mass, especially appendicular lean mass (ALM), was assessed by the two methods explained below. ALM was calculated as the sum of lean mass of arms and legs [12].
Bioelectrical Impedance Analysis (BIA)
A multi-frequency bioelectrical impedance analyzer, InBody S10 Biospace device (Biospacte Co, Ltd, Korea/Model JMW140) was used according to the manufacturer’s guidelines. BIA estimates body composition using the difference of conductivity of the various tissues due to the difference of their biological characteristics. Conductivity is proportional to water content, and more specifically to electrolytes, and it decreases as the cell approaches a perfect spherical shape. Adipose tissue is composed of round shaped cell and contains relatively little water compared to other tissues like muscle; therefore conductivity will decrease as body fat increases. In practice, electrodes are placed at 8 precise tactile-points of the body to achieve a multi-segmental frequency analysis. A total of 30 impedance measurements are obtained using 6 different frequencies (1 kHz, 5 kHz, 50 kHz, 250 kHz, 500 kHz, 1000 kHz) at the 5 following segments of the body: right and left arms, trunk, right and left legs.
Dual energy X-ray Absorptiometry (DXA)
As reference method, DXA scan (Hologic QDR Discovery device, Inc USA) was used for the measurement of whole and regional body composition, including a three-compartment model estimating body composition in terms of fat, bone mineral, and all other fat-free mass that does not include bone. DXA provides thus bone density estimates, and regional estimates of body composition (i.e. parts of the body), by measuring body’s absorbance of X-rays at two different energies using the fact that fat, bone mineral, and fat-free soft tissue have different absorption properties. The subjects were positioned for whole-body scans according to the manufacturer’s protocol. Subject laid in a supine position on the scanner table, with straight-legs and their arms close to the body. They were instructed to remain as still as possible for the duration of the scan. Whole-body composition analysis provided data on different regions of interest, including trunk, arms and legs. The DXA machine was calibrated daily against a phantom spine containing composites of bone, fat and lean tissue supplied by the manufacturer before testing. This procedure has been validated for general DXA use [13].
Collected data
Other variables were collected to characterize the population.
First, grip strength was assessed by means of a hydraulic dynamometer (Saehan Corporation, MSD Europe Bvba, Belgium, an isometric hydraulic hand dynamometer). According to the American Society of Hand Therapists, this particular instrument can provide the most stable results during repeated gripping trials. Its excellent test-retest reliability has been confirmed in many studies, with obtained IntraClass Correlation Coefficient (ICC) values ranging between 0.81 and 0.98 [14]. In our study, we used the following standardized protocol for the measurement of grip strength [15]. The participant was asked to sit comfortably on a standard chair with legs, back support and fixed arms. He was then advised to squeeze as hard as possible the hand dynamometer for up to six seconds and then relax. Three measurements for each hand, alternating sides, were performed consecutively and without rest. According to Watanabe [16], continuous measurements are not affected by fatigue, especially in the dominant hand. To encourage the subjects to get a score as high as possible, the best of the six grip strength measurements was recorded and later used in statistical analyses, as recently recommended by Roberts [15].
Thereafter, participant’s leisure time activity was evaluated using the short version of the Minnesota Leisure Time Physical Activity Questionnaire. This questionnaire asks participants about types, frequency and duration of their leisure time activity (average hour/day in the following four categories: walking, doing gymnastics or workouts, engaging in sports, and doing household activities). The kcal burned per day was calculated using the activity metabolic index, which allows the calories burned to be measured using the metabolic equivalent of tasks [17,18].
Statistical analysis
A Shapiro-Wilk test verified the normal distribution for all parameters. Quantitative variables were expressed by mean ± standard deviation (SD), or by median and interquartile range (P25-P75) for asymmetric distributions. Qualitative variables were expressed by number and percentage. Agreement between tools was assessed by means of the Bland Altman method and reliability by means of the IntraClass Correlation Coefficient (ICC) [14]. ICC was also computed to assess the reliability of the test-retest performed by the same operator or by two different ones. The closer the coefficient is to 1, the higher the reliability. We considered an ICC over 0.90 as very high, between 0.70 and 0.89 as high and between 0.50 and 0.69 as moderate [19]. Body composition obtained with the different methods (BIA and DXA) were compared using the t-test or Wilcoxon signed-rank test when appropriate. Analyzes were also performed by gender and by age category (i.e. 18–34 years; 35–64 years and 65+ years).
A multiple regression was conducted to obtain a muscle mass assessed by BIA, close to that measured by DXA. All calculations were performed by using Statistica 10 software, SAS statistical package (version 9.3 for windows) and R statistical packages. Results were considered to be statistically significant at the 5% critical level (p < 0.05).