Use of a Durnin-Womersley formula to estimate change in subcutaneous fat content in HIV-infected subjects

Background: HIV-infected individuals may develop malnutrition or lipodystrophy, leading to losses of subcutaneous adipose tissue (SAT). Objective: We compared the ability of a Durnin-Womersley formula for total adipose tissue (TAT) to estimate change in SAT with the use of whole-body magnetic resonance imaging (MRI) as a criterion measure. Design: We analyzed data from 2 clinical trials: a prospective randomized trial of protein supplements, progressive resistance training, or combined treatment in 29 malnourished, HIV-positive women, and a prospective open-label trial of recombinant human growth hormone in 25 HIV-infected subjects with visceral adipose tissue (VAT) accumulation. Changes in fat by the Durnin-Womersley formula and in SAT, TAT, and VAT by MRI were compared by linear regression, and Bland-Altman analyses were used to assess the agreement between the prediction and criterion methods. The repeatability of the Durnin-Womersley measurement was evaluated in 14 weight-stable, healthy adults studied twice within 1 y. Results: At baseline, Durnin-Womersley fat was significantly associated with SAT (r² = 0.75, P < 0.001) and TAT (r² = 0.79, P < 0.001) but not with VAT. Change in Durnin-Womersley fat was significantly associated with change in SAT (r² = 0.66, P < 0.001) and in TAT (r² = 0.57, P < 0.001) but not in VAT. The limits of agreement for the Durnin-Womersley estimation of change in SAT were -3.4 to 2.6 kg and the SEE was 1.5 kg. The SEE for repeated measures of SAT in healthy control subjects was 0.84. Conclusions: The Durnin-Womersley formula can be used to predict change in SAT. The limits of agreement and the SEE for predicting change in SAT by MRI are approximately twice as great as the error of repeated Durnin-Womersley measures in control subjects.