Body composition in early onset eating disorders

Abstract

Background: Body mass index (BMI) or equivalent weight for height indices are the most widely used measures of body composition in early onset and adolescent eating disorders. Although of value as screening instruments the limitation in disease states is their inability to discriminate fat and fat-free components of body weight. Objective: To compare height-adjusted fat and fat-free components of body composition in children and young adolescents with different types of eating disorders with those of age matched reference children. Design: Weight, height, triceps and subscapular skinfold thickness were measured in 172 children (aged 7 -16 y) with eating disorders receiving specialist treatment. Fat mass index (FMI) and fat-free mass index (FFMI) were calculated using Slaughter's and Deurenberg's equations and normalisation for height. Using data from 157 normal children, representative of the UK 1990 growth reference data, reference curves for FMI and FFMI AE 2 s.d. were derived. Results for patient groups were superimposed on these reference curves.Results: FMI and FFMI were both reduced in eating disorders associated with malnutrition, including anorexia nervosa (AN). AN subjects did not differ from other subjects with comparable degrees of malnutrition. Children with eating disorders of normal weight, such as bulimia nervosa and selective eating, did not differ significantly from reference children in their relative FM and FFM. Conclusions: FM and FFM merit independent consideration in disorders of malnutrition in children, rather than expressing data as percentage body fat or percentage BMI. The implications of loss of FFM on growth and development merit further investigation.

Thigh Circumference as a Useful Predictor of Body Fat in Adolescent Girls with Anorexia Nervosa

“…The most widely used weight-for-height indices, such as BMI, are considered clinical predictors of the nutritional status during growth; however, their inability to discriminate between fat and lean body components limits inferences that could be made [2]. In fact, body weight itself can be used as a single measure to predict body fat in healthy children, particularly in the absence of any additional methods [21].…”

“…Therefore, body weight and nutritional status in AN are considered major health problems in managing these patients because somatic condition, weight gain and restoration of appropriate body composition is of critical value in the process of cure. Body weight itself does not provide sufficient information about the nutritional status in this condition [1,2]. Although body weight is a commonly used criterion for the classification of the nutritional status in AN, body composition analysis can provide an additional valuable information beyond weight.…”

“…Several studies have reported the use of anthropometry in the assessment of nutritional status in AN patients, and most of them focused on comparisons between BMI, skinfold thickness measurements, underwater weighing and dual-energy X-ray absorptiometry (DXA) or other techniques (usually noninvasive) such as air displacement plethysmography, neutron activation analysis and bioelectrical impedance analysis [2,4,5,6,7]. A variety of equations have also been developed for the prediction of adiposity, however, the calculations provide only an approximate estimation of body fat.…”

Nutritional Evaluation of Eating Disorders

“…Nutritional requirements are directly related to the basal metabolic rate, which is often diminished in anorexia nervosa and augmented in bulimia nervosa Nicholls et al, 2002;Russell et al, 2001;Scalfi et al, 2010;Schebendach et al, 1997;Sedlet et al, 1989). However, anorectic patients tend to present high respiratory coefficient and induced thermogenesis, which may explain the high weight reduction in emaciated patients and the difficulties to recovery and maintain a healthy weight (Russell et al, 2001).…”

“…The limitation of the use of percentile is the acquisition of reference tables validated in the study population (Lama et al, 2002;Nicholls et al, 2002;Olveira et al, 2007). It should be taken as normal the percentiles of 5-95, while focusing on the 50 percentile (Lama et al, 2002).…”