Prediction of body tissue distribution with multi-dimensional anthropometry

Onderzoeksoutput: BookResearch

Samenvatting

Although newly developed techniques have contributed significantly to the knowledge in body composition over the last decades, their advent has not decreased the popularity of anthropometry which still remains the most widely used method for quantifying body composition today. Since the accepted 'gold' standards for regional BC estimation in vivo, such as MRI and CT are difficult to access in routine health care situations, it remains important to use anthropometry in the screening and diagnosis of health and nutritional risk. However, the validity of anthropometric variables for estimating regional body tissue distribution is largely unknown, especially with regard to the trunk region. Therefore, four studies were carried out in an attempt to contribute to the validation of techniques to estimate body composition. The general objective of this work is to determine the value and limitations of multi-dimensional anthropometry in estimating regional body composition for use in clinical practice.
In the first study, the relationship of the body mass index and the waist circumference with muscle/adipose tissue mass ratios and with trunk adipose tissue distribution was examined, based on an anatomical 5-compartment model obtained by dissection of cadavers of elderly persons. The results of this study indicate that a single measurement of body weight and height and/or waist circumference is of limited use for assessing body tissue distribution in elderly individuals due to significant between subject variability in body tissue distribution, especially in the intermediate body mass index ranges. Therefore it was concluded that the body mass index and the waist circumference lack discriminative power to differentiate between metabolic compartments in elderly persons.
The observation of the first study, that simple anthropometric indexes are insufficiently accurate to determine body tissue distribution, and the knowledge that regression equations may have additive predictive power, have prompted to verify whether regional body tissue distribution can be predicted accurately using a combination of multi-dimensional anthropometric variables. To this end two similar anthropometric modeling studies, each with individual objectives, were conducted. The first modeling study aimed at systematically developing and cross-validating regression-based anthropometric prediction equations for the accurate estimation of lean tissue distribution in all major body regions in a sample of young and middle-aged adults. In the second modeling study, the accuracy and concordance of anthropometrically-derived prediction equations for the estimation of regional fat mass distribution, including the ratio of central to peripheral fat, was explored. The findings of the first study suggested that segmental lean mass can be predicted with consistent accuracy in groups of young
adults using multi-dimensional anthropometry, confirming its potential in assessing regional lean tissue distribution. However, when using these prediction formulae in clinical practice, caution is warranted since estimation errors of 8% to 14% may occur in certain individuals depending on the anatomical region of interest. The results of the second study indicated that fat mass distribution can be predicted accurately in young and middle-aged adults without systematic differences with the criterion method, suggesting method interchangeability. This study also showed that defining fat mass distribution thresholds, for example a fat mass ratio of 2.0 standard deviations or more above a reference fat mass ratio of young adults, by analogy with osteoporosis assessment, might prove itself a useful tool to detect abnormal fat distribution. Nevertheless, the clinician needs to keep in mind that misclassification may still occur in a minority of subjects, in particular in those with borderline values. Notwithstanding this limitation the use of multidimensional anthropometry will likely result in fewer 'misclassifications' than either waist-to-hip ratio or other simple indices of body fat distribution.
Based on the above-described conclusions, that both regional lean and fat mass distribution can be predicted fairly accurately using a combination of multiple anthropometric variables, the question was raised whether the criterion method used in the modeling studies, dual energy X-ray absorptiometry, is a suitable assessment tool for estimating relevant metabolic tissue compartments. Since dual energy X-ray absorptiometry cannot distinguish between different adipose tissue and adipose tissue free sub-depots, it remains unclear how the assumed inter-individual variability in (regional) adipose tissue and muscle distribution is translated to dual energy X-ray absorptiometry output. Therefore, the last objective of this work was to compare and relate regional dual energy X-ray absorptiometry variables with absolute tissue masses obtained by computer tomography scanning in legs of cadavers of elderly persons. The findings of this validation study confirm that dual energy X-ray absorptiometry variables and absolute tissue masses are significantly related. However, it is not recommended to use these body composition variables interchangeably as they represent different quantitative and physiological body components. This observation highlights the importance of a precise definition of terms and organizational rules used in BC research. Given the significant variability in tissue distribution between and within genders, dual energy X-ray absorptiometry output should be interpreted differently in elderly male and female persons.
Originele taal-2English
UitgeverijUnknown
Aantal pagina's103
UitgavePhD thesis
StatusPublished - 2013

Vingerafdruk

Duik in de onderzoeksthema's van 'Prediction of body tissue distribution with multi-dimensional anthropometry'. Samen vormen ze een unieke vingerafdruk.

Citeer dit