Background:  The diagnosis of lipoatrophy is currently made by patient self-report and physicians’ clinical assessment. This does not include dual energy X-ray absorptiometry (DEXA) scans or fat biopsies evaluation of mitochondrial parameters. In this study, we assessed the correlation between clinical assessment of lipoatrophy and DEXA measured limb fat and fat mtDNA levels.

Methods:  From a cross-sectional cohort, we assessed 20 HIV⁺ subjects for metabolic parameters. Evaluations included excisional subcutaneous fat biopsies from the abdomen for measurement of mtDNA copies/cell by real time polymerase chain reaction (PCR), total body DEXA scanning, fasting homeostasis model assessment (HOMA), and lipid panel. A body image questionnaire was filled separately and independently by patients and physician. The questionnaire asked for a rating of fat loss in predefined body areas:  arms, legs, buttocks, and face. Assessments within each of these sites were rated as:  0 = absent, 1 = mild, 2 = moderate, and 3 = severe; the lipoatrophy score could vary between 0 and 12. The relationship between continuous variables was reported using Spearman’s correlation

Results:  Of the 20 patients enrolled (15 males, 11 white, age 42 years), 17 were treated with thymidine analogue-containing regimen (8 stavudine [d4T] and 9 zidovudine [ZDV]) and 3 were naïve to all ART; 6 were receiving a protease inhibitor. Median (range) lipoatrophy score generated by patients and physician were 6 (0 to 12) and 8.5 (0 to 12), respectively. Median limb fat, fat mtDNA, and peripheral blood mononuclear cell (PBMC) mtDNA were 4.43 kg, 1011, and 387.5 copies/cell, respectively. There was a strong positive correlation between the patient-generated lipoatrophy score and the physician score (r = 0.80; p <0.0001). In addition, there was a strong negative correlation between DEXA-measured limb fat and lipoatrophy scores generated either by the patients or the physician (r = –0.58; p = 0.0075 and r = –0.59; p = 0.0058, respectively), and between fat mtDNA levels and lipoatrophy scores generated by the patients or the physician (r = –0.51; p = 0.02 and r = –0.47; p = 0.03, respectively). Additionally, fat mtDNA tended to correlate with leg fat (r = 0.42; p = 0.06), but not with arm fat (r = 0.13; p = 0.58). There was no correlation between PBMC and fat mtDNA levels (r = –0.38; p = 0.10), nor between PBMC mtDNA and limb fat (r = –0.14; p = 0.56).

Conclusions:  The clinical assessment of lipoatrophy by independently generated patients’ and physician’ scores strongly correlated with DEXA-measured limb fat and with subcutaneous fat mtDNA levels.