Long-Term Body Fat Outcomes in Antiretroviral-Naive Participants Randomized to Nelfinavir or Efavirenz or Both Plus Dual Nucleosides

BACKGROUND: Long-term regional body fat outcomes have not been well described in randomized antiretroviral drug trials. METHODS: Dual x-ray absorptiometry (DXA) scans were performed every 16 weeks on a subset of 157 antiretroviral-naive participants who were randomized to receive nelfinavir, efavirenz, or both, combined with zidovudine and lamivudine or stavudine and didanosine, in a multicenter trial. Participants with any available data after baseline up to week 144 contributed to this as-treated analysis. RESULTS: Limb fat increased similarly in all groups during the first 32 weeks. After week 32, limb fat changed by - 1.7% per year for zidovudine-lamivudine and by -19.0% per year for didanosine-stavudine (mixed model analysis of variance [MMANOVA], P < 0.0001). Adjusting for nucleoside backbone, there was an additional decrease in limb fat of -8.7% per year for the combined nelfinavir and nelfinavir + efavirenz group compared with the efavirenz group (MMANOVA, P = 0.03). Among participants receiving zidovudine-lamivudine, after week 32, limb fat changed by +2.7% per year with efavirenz and by -7.9% per year for the combined nelfinavir and nelfinavir + efavirenz group (MMANOVA, P = 0.03). CONCLUSIONS: Over 144 weeks, zidovudine-lamivudine was superior to didanosine-stavudine with regard to limb fat loss. The combination of zidovudine, lamivudine, and efavirenz showed no overall pattern suggesting limb fat loss over time and was significantly superior to the pooled zidovudine-lamivudine-nelfinavir (with and without efavirenz) arms.