A radiographic method of quantifying protein-calorie undernutrition

Estimation of midarm adipose tissue and muscle by the anthropometric technique is based on the idealized assumption that the arm and its muscle compartments are circular, and that fat is distributed evenly around the arm. We examined the validity of these assumptions by computerized axial tomography of the midarm in 21 subjects ranging from 65 to 255% of ideal body weight. Computerized axial tomography identified three errors inherent in the anthropometric method: 1) The arm and its muscle compartment were rarely circular, but resembled instead an ellipse and "cloverleaf", respectively; 2) fat was distributed asymmetrically around the arm, and furthermore when triceps skinfold was less than 5 mm, no fat was radiographically detectable, and 3) muscle are calculated by the anthropometric method includes bone area. Since bone area was not influenced by nutritional status, anthropometric "muscle area" underestimated the degree of muscle atrophy in undernutrition. Despite these limitations, in subjects 60 to 120% of ideal body weight anthropometric estimates of midarm muscle area (MAMA) and fat area did not differ greatly from the radiographic values. Anthropometric MAMA was consistently greater than the radiographic value by 15 to 25%, while midarm fat area agreed within +/- 10%. The error in the anthropometric MAMA could be nearly eliminated by two types of correction: expressing MAMA as a percentage of normal, and correcting for bone content by subtracting midarm bone area (6.3 and 4.7 cm2 for men and women). In subjects greater than 150% ideal body weight, however, anthropometric estimates of MAMA and midarm fat area differed from the radiographic values by greater than 50% even after the above two types of correction. Midarm computerized axial tomography scan provides an accurate alternative to the anthropometric method for estimating midarm muscle and fat in these obese individuals.