Background: The relationship between the subjective diagnosis of lipoatrophy and the objective amount of limb fat loss in HIV-infected adults is unclear.
Methods: Using Medline, we identified published articles reporting the amount of arm, leg or limb fat measured by dual X-ray absorptiometry in HIV-infected patients with moderate-severe lipoatrophy and in healthy non-HIV-infected adults. We calculated the relative content of fat in the limbs, arms and legs of lipoatrophic patients with regard to the weighted arithmetic means of those fat values in healthy controls.
Results: We found 799 patients from 10 articles, and 73 healthy controls from two articles. Limb fat ranged from 2.6 to 4.4 kg in patients, and from 7.1 to 7.2 kg in controls. Both patients and controls were almost exclusively men, of white race, and in their forties. Weighted arithmetic means of arm, leg and limb fat in HIV-infected patients with clinically evident lipoatrophy were 1.0, 2.1 and 3.1 kg, respectively (48, 41 and 43% relative to healthy non-HIV-infected males, respectively).
Conclusions: The diagnosis of lipoatrophy was highly correlated with the amount of limb fat, irrespective of the investigators. HIV-infected men with clinically evident lipoatrophy had a limb fat loss of >50% compared with non-HIV-infected healthy males.