HAART increases body fat at first; both d4T and AZT lead to subsequent fat loss

Simon Collins, HIV i-Base

Evaluation of the contribution of individual drugs and HAART itself to the varied manifestations of the lipodystrophy syndrome was hampered in early studies by poor monitoring, limited duration and subjective and retrospective reporting. In addition to linking peripheral fat loss to both d4T and AZT-containing regimens and highlighting an increase in fat over the initial months of treatment that is not generally appreciated by patients concerned about the risk of lipodystrophy, important observations on blood lipids and bone mineral density were also reported in this Australian study.

This prospective study, published in 2 May issue of AIDS, recruited 40 treatment naïve men (median age 39, IQR 30-49) between July 1997 and May 2000, monitoring them at baseline, and 12 and 24 weeks after starting treatment and every 24 weeks thereafter. Median CD4 and viral load were 246 cells/mm3 (IQR 62-430 cells/mm3) and 5 log copies/ml (IQR 4.3-5.7 copies/ml).

The primary hypothesis was that lipoatrophy occurs as a result of HIV therapy, with primary endpoint change of limb fat measured by DEXA. Secondary endpoint included change in abdominal fat and other metabolic parameters. Choice of treatment was non-randomised and left to the individual doctor and patient. In addition to the routine parameters of CD4, viral load, fasting total cholesterol (TC), LDL and HDL cholesterol, triglycerides (TG), glucose and insulin, DEXA scans from a single site were used to assess changes in central abdominal fat (CAF), limb fat (LF) and lean mass (LM), and spinal bone mineral density.

At the time of analysis 90% and 50% of patients had been followed for 96 and 144 weeks respectively. All median baseline parameters were within normal limits. Almost half the group started treatment with a PI, five using PI+NNRTI-based treatment. Dual nucleoside background was equally distributed between ddI+d4T, d4T+3TC or AZT+3TC.

40% of men changed treatment over the course of the study, nearly all related to toxicity management, none related to lipodystrophy. Most changes did not occur during the first months of treatment although 11 changes from d4T occurred after a mean of seven months (three-12 months).

The following key results relating to body changes were reported:

  • BMI increased significantly (22.5 to 24, P<0.05) and lean mass increased by 3% (p<0.01) by week 12, which was sustained to week 48, and returned to baseline by week 96.
  • Total body fat increased by median 20% by week 24 (p<0.001) but was significantly higher in PI vs NNRTI-based therapy, and remained stable thereafter.
  • Limb fat (5.6 vs 6.9kg) central fat (1.0 vs 1.3kg) and lean mass (52.5 vs 54.7kg) all increased over first 24 weeks (coinciding with greatest viral load and CD4 changes).
  • From week 24 there was a selective and progressive loss of limb fat [median 13% (QR 0.9-26.3) loss per year] with d4T most significantly associated with increased rate of loss in multivariate analysis.
  • Central abdominal fat was significantly higher in patients using PI- compared to NNRTI-based combinations. Increases in CAF at week 24 were maintained after week 24 and remained significantly higher than baseline at week 144.
  • Hypercholesterolaemia developed early in treatment, whereas hypertriglyceridaemia, hyperinsulinaemia and decreased bone mineral density developed later.
  • Bone mineral density T-score rose slightly to week 24 and fell significantly to week 48, maintained over follow up. Percentage of people with osteopenia (T-score <1.0) increased from 13% at baseline to 22% at week 144, although only one additional score consistent with osteoporosis (T-score < – 2.5) was noted during the study.
  • The largest changes in CD4 cell counts and HIV viral load, seen early into treatment, were associated with gain rather than loss of fat.

The authors conclude that this is the first prospective study demonstrating that treatment with antiretrovirals results in progressive, selective loss of limb fat. Also that the loss of limb fat occurred after the period of most intense immune restoration, making an immune aetiology unlikely.


Mallon PWG, Miller J, Cooper DA, Carr A. Prospective evaluation of the effects of antiretroviral therapy on body composition in HIV-1-infected men starting therapy. AIDS 2003; 17(7):971-979


Because the study first started recruiting in 1997 only shortly after combination therapy became widely available, it is likely to include a number of patients with more advanced HIV disease. This may explain some of the early increases in weight and lean mass – but similar results have been reported in a recent ACTG study.

However the link of both d4T and AZT treatment to lipoatrophy is very important, and with availability of a wide range of nucleosides for initial therapy, many doctors are choosing not to use either drug in first line therapy. This has been reflected in the recent decision not to recommend d4T for initial therapy in the BHIVA guidelines.

The report of bone mineral changes over such a short period again highlights the potential role for baseline DEXA scan prior to treatment.

Links to other websites are current at date of posting but not maintained.